diff options
Diffstat (limited to 'drivers/block')
53 files changed, 2383 insertions, 1914 deletions
diff --git a/drivers/block/DAC960.c b/drivers/block/DAC960.c index ce1fa923c414..c5f22bb0a48e 100644 --- a/drivers/block/DAC960.c +++ b/drivers/block/DAC960.c @@ -2533,9 +2533,8 @@ static bool DAC960_RegisterBlockDevice(DAC960_Controller_T *Controller) Controller->RequestQueue[n] = RequestQueue; blk_queue_bounce_limit(RequestQueue, Controller->BounceBufferLimit); RequestQueue->queuedata = Controller; - blk_queue_max_hw_segments(RequestQueue, Controller->DriverScatterGatherLimit); - blk_queue_max_phys_segments(RequestQueue, Controller->DriverScatterGatherLimit); - blk_queue_max_sectors(RequestQueue, Controller->MaxBlocksPerCommand); + blk_queue_max_segments(RequestQueue, Controller->DriverScatterGatherLimit); + blk_queue_max_hw_sectors(RequestQueue, Controller->MaxBlocksPerCommand); disk->queue = RequestQueue; sprintf(disk->disk_name, "rd/c%dd%d", Controller->ControllerNumber, n); disk->major = MajorNumber; @@ -7134,7 +7133,7 @@ static struct DAC960_privdata DAC960_P_privdata = { .MemoryWindowSize = DAC960_PD_RegisterWindowSize, }; -static struct pci_device_id DAC960_id_table[] = { +static const struct pci_device_id DAC960_id_table[] = { { .vendor = PCI_VENDOR_ID_MYLEX, .device = PCI_DEVICE_ID_MYLEX_DAC960_GEM, diff --git a/drivers/block/Kconfig b/drivers/block/Kconfig index 77bfce52e9ca..de277689da61 100644 --- a/drivers/block/Kconfig +++ b/drivers/block/Kconfig @@ -76,6 +76,17 @@ config BLK_DEV_XD It's pretty unlikely that you have one of these: say N. +config GDROM + tristate "SEGA Dreamcast GD-ROM drive" + depends on SH_DREAMCAST + help + A standard SEGA Dreamcast comes with a modified CD ROM drive called a + "GD-ROM" by SEGA to signify it is capable of reading special disks + with up to 1 GB of data. This drive will also read standard CD ROM + disks. Select this option to access any disks in your GD ROM drive. + Most users will want to say "Y" here. + You can also build this as a module which will be called gdrom. + config PARIDE tristate "Parallel port IDE device support" depends on PARPORT_PC @@ -103,17 +114,6 @@ config PARIDE "MicroSolutions backpack protocol", "DataStor Commuter protocol" etc.). -config GDROM - tristate "SEGA Dreamcast GD-ROM drive" - depends on SH_DREAMCAST - help - A standard SEGA Dreamcast comes with a modified CD ROM drive called a - "GD-ROM" by SEGA to signify it is capable of reading special disks - with up to 1 GB of data. This drive will also read standard CD ROM - disks. Select this option to access any disks in your GD ROM drive. - Most users will want to say "Y" here. - You can also build this as a module which will be called gdrom. - source "drivers/block/paride/Kconfig" config BLK_CPQ_DA diff --git a/drivers/block/amiflop.c b/drivers/block/amiflop.c index 055225839024..832798aa14f6 100644 --- a/drivers/block/amiflop.c +++ b/drivers/block/amiflop.c @@ -54,6 +54,7 @@ */ #include <linux/module.h> +#include <linux/slab.h> #include <linux/fd.h> #include <linux/hdreg.h> @@ -65,6 +66,7 @@ #include <linux/blkdev.h> #include <linux/elevator.h> #include <linux/interrupt.h> +#include <linux/platform_device.h> #include <asm/setup.h> #include <asm/uaccess.h> @@ -1695,34 +1697,18 @@ static struct kobject *floppy_find(dev_t dev, int *part, void *data) return get_disk(unit[drive].gendisk); } -static int __init amiga_floppy_init(void) +static int __init amiga_floppy_probe(struct platform_device *pdev) { int i, ret; - if (!MACH_IS_AMIGA) - return -ENODEV; - - if (!AMIGAHW_PRESENT(AMI_FLOPPY)) - return -ENODEV; - if (register_blkdev(FLOPPY_MAJOR,"fd")) return -EBUSY; - /* - * We request DSKPTR, DSKLEN and DSKDATA only, because the other - * floppy registers are too spreaded over the custom register space - */ - ret = -EBUSY; - if (!request_mem_region(CUSTOM_PHYSADDR+0x20, 8, "amiflop [Paula]")) { - printk("fd: cannot get floppy registers\n"); - goto out_blkdev; - } - ret = -ENOMEM; if ((raw_buf = (char *)amiga_chip_alloc (RAW_BUF_SIZE, "Floppy")) == NULL) { printk("fd: cannot get chip mem buffer\n"); - goto out_memregion; + goto out_blkdev; } ret = -EBUSY; @@ -1791,18 +1777,13 @@ out_irq2: free_irq(IRQ_AMIGA_DSKBLK, NULL); out_irq: amiga_chip_free(raw_buf); -out_memregion: - release_mem_region(CUSTOM_PHYSADDR+0x20, 8); out_blkdev: unregister_blkdev(FLOPPY_MAJOR,"fd"); return ret; } -module_init(amiga_floppy_init); -#ifdef MODULE - #if 0 /* not safe to unload */ -void cleanup_module(void) +static int __exit amiga_floppy_remove(struct platform_device *pdev) { int i; @@ -1819,12 +1800,25 @@ void cleanup_module(void) custom.dmacon = DMAF_DISK; /* disable DMA */ amiga_chip_free(raw_buf); blk_cleanup_queue(floppy_queue); - release_mem_region(CUSTOM_PHYSADDR+0x20, 8); unregister_blkdev(FLOPPY_MAJOR, "fd"); } #endif -#else +static struct platform_driver amiga_floppy_driver = { + .driver = { + .name = "amiga-floppy", + .owner = THIS_MODULE, + }, +}; + +static int __init amiga_floppy_init(void) +{ + return platform_driver_probe(&amiga_floppy_driver, amiga_floppy_probe); +} + +module_init(amiga_floppy_init); + +#ifndef MODULE static int __init amiga_floppy_setup (char *str) { int n; @@ -1839,3 +1833,5 @@ static int __init amiga_floppy_setup (char *str) __setup("floppy=", amiga_floppy_setup); #endif + +MODULE_ALIAS("platform:amiga-floppy"); diff --git a/drivers/block/aoe/aoeblk.c b/drivers/block/aoe/aoeblk.c index 3af97d4da2db..035cefe4045a 100644 --- a/drivers/block/aoe/aoeblk.c +++ b/drivers/block/aoe/aoeblk.c @@ -9,6 +9,7 @@ #include <linux/backing-dev.h> #include <linux/fs.h> #include <linux/ioctl.h> +#include <linux/slab.h> #include <linux/genhd.h> #include <linux/netdevice.h> #include "aoe.h" diff --git a/drivers/block/aoe/aoechr.c b/drivers/block/aoe/aoechr.c index 62141ec09a22..4a1b9e7464aa 100644 --- a/drivers/block/aoe/aoechr.c +++ b/drivers/block/aoe/aoechr.c @@ -8,6 +8,7 @@ #include <linux/blkdev.h> #include <linux/completion.h> #include <linux/delay.h> +#include <linux/slab.h> #include <linux/smp_lock.h> #include <linux/skbuff.h> #include "aoe.h" diff --git a/drivers/block/aoe/aoecmd.c b/drivers/block/aoe/aoecmd.c index 64a223b0cc22..5674bd01d96d 100644 --- a/drivers/block/aoe/aoecmd.c +++ b/drivers/block/aoe/aoecmd.c @@ -5,6 +5,7 @@ */ #include <linux/ata.h> +#include <linux/slab.h> #include <linux/hdreg.h> #include <linux/blkdev.h> #include <linux/skbuff.h> diff --git a/drivers/block/aoe/aoedev.c b/drivers/block/aoe/aoedev.c index fa67027789aa..0849280bfc1c 100644 --- a/drivers/block/aoe/aoedev.c +++ b/drivers/block/aoe/aoedev.c @@ -8,6 +8,7 @@ #include <linux/blkdev.h> #include <linux/netdevice.h> #include <linux/delay.h> +#include <linux/slab.h> #include "aoe.h" static void dummy_timer(ulong); diff --git a/drivers/block/aoe/aoenet.c b/drivers/block/aoe/aoenet.c index ce0d62cd71b2..4d3bc0d49df5 100644 --- a/drivers/block/aoe/aoenet.c +++ b/drivers/block/aoe/aoenet.c @@ -4,6 +4,7 @@ * Ethernet portion of AoE driver */ +#include <linux/gfp.h> #include <linux/hdreg.h> #include <linux/blkdev.h> #include <linux/netdevice.h> diff --git a/drivers/block/ataflop.c b/drivers/block/ataflop.c index a5af1d6dda8b..e35cf59cbfde 100644 --- a/drivers/block/ataflop.c +++ b/drivers/block/ataflop.c @@ -1470,8 +1470,6 @@ repeat: void do_fd_request(struct request_queue * q) { - unsigned long flags; - DPRINT(("do_fd_request for pid %d\n",current->pid)); while( fdc_busy ) sleep_on( &fdc_wait ); fdc_busy = 1; diff --git a/drivers/block/brd.c b/drivers/block/brd.c index 4f688434daf1..f1bf79d9bc0a 100644 --- a/drivers/block/brd.c +++ b/drivers/block/brd.c @@ -15,9 +15,9 @@ #include <linux/blkdev.h> #include <linux/bio.h> #include <linux/highmem.h> -#include <linux/gfp.h> #include <linux/radix-tree.h> #include <linux/buffer_head.h> /* invalidate_bh_lrus() */ +#include <linux/slab.h> #include <asm/uaccess.h> @@ -133,6 +133,28 @@ static struct page *brd_insert_page(struct brd_device *brd, sector_t sector) return page; } +static void brd_free_page(struct brd_device *brd, sector_t sector) +{ + struct page *page; + pgoff_t idx; + + spin_lock(&brd->brd_lock); + idx = sector >> PAGE_SECTORS_SHIFT; + page = radix_tree_delete(&brd->brd_pages, idx); + spin_unlock(&brd->brd_lock); + if (page) + __free_page(page); +} + +static void brd_zero_page(struct brd_device *brd, sector_t sector) +{ + struct page *page; + + page = brd_lookup_page(brd, sector); + if (page) + clear_highpage(page); +} + /* * Free all backing store pages and radix tree. This must only be called when * there are no other users of the device. @@ -189,6 +211,24 @@ static int copy_to_brd_setup(struct brd_device *brd, sector_t sector, size_t n) return 0; } +static void discard_from_brd(struct brd_device *brd, + sector_t sector, size_t n) +{ + while (n >= PAGE_SIZE) { + /* + * Don't want to actually discard pages here because + * re-allocating the pages can result in writeback + * deadlocks under heavy load. + */ + if (0) + brd_free_page(brd, sector); + else + brd_zero_page(brd, sector); + sector += PAGE_SIZE >> SECTOR_SHIFT; + n -= PAGE_SIZE; + } +} + /* * Copy n bytes from src to the brd starting at sector. Does not sleep. */ @@ -300,6 +340,12 @@ static int brd_make_request(struct request_queue *q, struct bio *bio) get_capacity(bdev->bd_disk)) goto out; + if (unlikely(bio_rw_flagged(bio, BIO_RW_DISCARD))) { + err = 0; + discard_from_brd(brd, sector, bio->bi_size); + goto out; + } + rw = bio_rw(bio); if (rw == READA) rw = READ; @@ -320,7 +366,7 @@ out: } #ifdef CONFIG_BLK_DEV_XIP -static int brd_direct_access (struct block_device *bdev, sector_t sector, +static int brd_direct_access(struct block_device *bdev, sector_t sector, void **kaddr, unsigned long *pfn) { struct brd_device *brd = bdev->bd_disk->private_data; @@ -434,9 +480,14 @@ static struct brd_device *brd_alloc(int i) goto out_free_dev; blk_queue_make_request(brd->brd_queue, brd_make_request); blk_queue_ordered(brd->brd_queue, QUEUE_ORDERED_TAG, NULL); - blk_queue_max_sectors(brd->brd_queue, 1024); + blk_queue_max_hw_sectors(brd->brd_queue, 1024); blk_queue_bounce_limit(brd->brd_queue, BLK_BOUNCE_ANY); + brd->brd_queue->limits.discard_granularity = PAGE_SIZE; + brd->brd_queue->limits.max_discard_sectors = UINT_MAX; + brd->brd_queue->limits.discard_zeroes_data = 1; + queue_flag_set_unlocked(QUEUE_FLAG_DISCARD, brd->brd_queue); + disk = brd->brd_disk = alloc_disk(1 << part_shift); if (!disk) goto out_free_queue; diff --git a/drivers/block/cciss.c b/drivers/block/cciss.c index 9291614ac6b7..51ceaee98f9f 100644 --- a/drivers/block/cciss.c +++ b/drivers/block/cciss.c @@ -257,6 +257,79 @@ static inline void removeQ(CommandList_struct *c) hlist_del_init(&c->list); } +static void cciss_free_sg_chain_blocks(SGDescriptor_struct **cmd_sg_list, + int nr_cmds) +{ + int i; + + if (!cmd_sg_list) + return; + for (i = 0; i < nr_cmds; i++) { + kfree(cmd_sg_list[i]); + cmd_sg_list[i] = NULL; + } + kfree(cmd_sg_list); +} + +static SGDescriptor_struct **cciss_allocate_sg_chain_blocks( + ctlr_info_t *h, int chainsize, int nr_cmds) +{ + int j; + SGDescriptor_struct **cmd_sg_list; + + if (chainsize <= 0) + return NULL; + + cmd_sg_list = kmalloc(sizeof(*cmd_sg_list) * nr_cmds, GFP_KERNEL); + if (!cmd_sg_list) + return NULL; + + /* Build up chain blocks for each command */ + for (j = 0; j < nr_cmds; j++) { + /* Need a block of chainsized s/g elements. */ + cmd_sg_list[j] = kmalloc((chainsize * + sizeof(*cmd_sg_list[j])), GFP_KERNEL); + if (!cmd_sg_list[j]) { + dev_err(&h->pdev->dev, "Cannot get memory " + "for s/g chains.\n"); + goto clean; + } + } + return cmd_sg_list; +clean: + cciss_free_sg_chain_blocks(cmd_sg_list, nr_cmds); + return NULL; +} + +static void cciss_unmap_sg_chain_block(ctlr_info_t *h, CommandList_struct *c) +{ + SGDescriptor_struct *chain_sg; + u64bit temp64; + + if (c->Header.SGTotal <= h->max_cmd_sgentries) + return; + + chain_sg = &c->SG[h->max_cmd_sgentries - 1]; + temp64.val32.lower = chain_sg->Addr.lower; + temp64.val32.upper = chain_sg->Addr.upper; + pci_unmap_single(h->pdev, temp64.val, chain_sg->Len, PCI_DMA_TODEVICE); +} + +static void cciss_map_sg_chain_block(ctlr_info_t *h, CommandList_struct *c, + SGDescriptor_struct *chain_block, int len) +{ + SGDescriptor_struct *chain_sg; + u64bit temp64; + + chain_sg = &c->SG[h->max_cmd_sgentries - 1]; + chain_sg->Ext = CCISS_SG_CHAIN; + chain_sg->Len = len; + temp64.val = pci_map_single(h->pdev, chain_block, len, + PCI_DMA_TODEVICE); + chain_sg->Addr.lower = temp64.val32.lower; + chain_sg->Addr.upper = temp64.val32.upper; +} + #include "cciss_scsi.c" /* For SCSI tape support */ static const char *raid_label[] = { "0", "4", "1(1+0)", "5", "5+1", "ADG", @@ -1344,26 +1417,27 @@ static int cciss_ioctl(struct block_device *bdev, fmode_t mode, kfree(buff); return -ENOMEM; } - // Fill in the command type + /* Fill in the command type */ c->cmd_type = CMD_IOCTL_PEND; - // Fill in Command Header - c->Header.ReplyQueue = 0; // unused in simple mode - if (iocommand.buf_size > 0) // buffer to fill + /* Fill in Command Header */ + c->Header.ReplyQueue = 0; /* unused in simple mode */ + if (iocommand.buf_size > 0) /* buffer to fill */ { c->Header.SGList = 1; c->Header.SGTotal = 1; - } else // no buffers to fill + } else /* no buffers to fill */ { c->Header.SGList = 0; c->Header.SGTotal = 0; } c->Header.LUN = iocommand.LUN_info; - c->Header.Tag.lower = c->busaddr; // use the kernel address the cmd block for tag + /* use the kernel address the cmd block for tag */ + c->Header.Tag.lower = c->busaddr; - // Fill in Request block + /* Fill in Request block */ c->Request = iocommand.Request; - // Fill in the scatter gather information + /* Fill in the scatter gather information */ if (iocommand.buf_size > 0) { temp64.val = pci_map_single(host->pdev, buff, iocommand.buf_size, @@ -1371,7 +1445,7 @@ static int cciss_ioctl(struct block_device *bdev, fmode_t mode, c->SG[0].Addr.lower = temp64.val32.lower; c->SG[0].Addr.upper = temp64.val32.upper; c->SG[0].Len = iocommand.buf_size; - c->SG[0].Ext = 0; // we are not chaining + c->SG[0].Ext = 0; /* we are not chaining */ } c->waiting = &wait; @@ -1514,7 +1588,6 @@ static int cciss_ioctl(struct block_device *bdev, fmode_t mode, c->Request = ioc->Request; if (ioc->buf_size > 0) { - int i; for (i = 0; i < sg_used; i++) { temp64.val = pci_map_single(host->pdev, buff[i], @@ -1670,14 +1743,9 @@ static void cciss_softirq_done(struct request *rq) /* unmap the DMA mapping for all the scatter gather elements */ for (i = 0; i < cmd->Header.SGList; i++) { if (curr_sg[sg_index].Ext == CCISS_SG_CHAIN) { - temp64.val32.lower = cmd->SG[i].Addr.lower; - temp64.val32.upper = cmd->SG[i].Addr.upper; - pci_dma_sync_single_for_cpu(h->pdev, temp64.val, - cmd->SG[i].Len, ddir); - pci_unmap_single(h->pdev, temp64.val, - cmd->SG[i].Len, ddir); + cciss_unmap_sg_chain_block(h, cmd); /* Point to the next block */ - curr_sg = h->cmd_sg_list[cmd->cmdindex]->sgchain; + curr_sg = h->cmd_sg_list[cmd->cmdindex]; sg_index = 0; } temp64.val32.lower = curr_sg[sg_index].Addr.lower; @@ -1796,12 +1864,9 @@ static int cciss_add_disk(ctlr_info_t *h, struct gendisk *disk, blk_queue_bounce_limit(disk->queue, h->pdev->dma_mask); /* This is a hardware imposed limit. */ - blk_queue_max_hw_segments(disk->queue, h->maxsgentries); - - /* This is a limit in the driver and could be eliminated. */ - blk_queue_max_phys_segments(disk->queue, h->maxsgentries); + blk_queue_max_segments(disk->queue, h->maxsgentries); - blk_queue_max_sectors(disk->queue, h->cciss_max_sectors); + blk_queue_max_hw_sectors(disk->queue, h->cciss_max_sectors); blk_queue_softirq_done(disk->queue, cciss_softirq_done); @@ -2368,7 +2433,7 @@ static int deregister_disk(ctlr_info_t *h, int drv_index, /* if it was the last disk, find the new hightest lun */ if (clear_all && recalculate_highest_lun) { - int i, newhighest = -1; + int newhighest = -1; for (i = 0; i <= h->highest_lun; i++) { /* if the disk has size > 0, it is available */ if (h->drv[i] && h->drv[i]->heads) @@ -2425,7 +2490,7 @@ static int fill_cmd(CommandList_struct *c, __u8 cmd, int ctlr, void *buff, c->Request.Type.Direction = XFER_READ; c->Request.Timeout = 0; c->Request.CDB[0] = cmd; - c->Request.CDB[6] = (size >> 24) & 0xFF; //MSB + c->Request.CDB[6] = (size >> 24) & 0xFF; /* MSB */ c->Request.CDB[7] = (size >> 16) & 0xFF; c->Request.CDB[8] = (size >> 8) & 0xFF; c->Request.CDB[9] = size & 0xFF; @@ -2694,7 +2759,7 @@ static void cciss_geometry_inquiry(int ctlr, int logvol, "cciss: reading geometry failed, volume " "does not support reading geometry\n"); drv->heads = 255; - drv->sectors = 32; // Sectors per track + drv->sectors = 32; /* Sectors per track */ drv->cylinders = total_size + 1; drv->raid_level = RAID_UNKNOWN; } else { @@ -3082,7 +3147,6 @@ static void do_cciss_request(struct request_queue *q) SGDescriptor_struct *curr_sg; drive_info_struct *drv; int i, dir; - int nseg = 0; int sg_index = 0; int chained = 0; @@ -3112,19 +3176,19 @@ static void do_cciss_request(struct request_queue *q) /* fill in the request */ drv = creq->rq_disk->private_data; - c->Header.ReplyQueue = 0; // unused in simple mode + c->Header.ReplyQueue = 0; /* unused in simple mode */ /* got command from pool, so use the command block index instead */ /* for direct lookups. */ /* The first 2 bits are reserved for controller error reporting. */ c->Header.Tag.lower = (c->cmdindex << 3); c->Header.Tag.lower |= 0x04; /* flag for direct lookup. */ memcpy(&c->Header.LUN, drv->LunID, sizeof(drv->LunID)); - c->Request.CDBLen = 10; // 12 byte commands not in FW yet; - c->Request.Type.Type = TYPE_CMD; // It is a command. + c->Request.CDBLen = 10; /* 12 byte commands not in FW yet; */ + c->Request.Type.Type = TYPE_CMD; /* It is a command. */ c->Request.Type.Attribute = ATTR_SIMPLE; c->Request.Type.Direction = (rq_data_dir(creq) == READ) ? XFER_READ : XFER_WRITE; - c->Request.Timeout = 0; // Don't time out + c->Request.Timeout = 0; /* Don't time out */ c->Request.CDB[0] = (rq_data_dir(creq) == READ) ? h->cciss_read : h->cciss_write; start_blk = blk_rq_pos(creq); @@ -3149,13 +3213,8 @@ static void do_cciss_request(struct request_queue *q) for (i = 0; i < seg; i++) { if (((sg_index+1) == (h->max_cmd_sgentries)) && !chained && ((seg - i) > 1)) { - nseg = seg - i; - curr_sg[sg_index].Len = (nseg) * - sizeof(SGDescriptor_struct); - curr_sg[sg_index].Ext = CCISS_SG_CHAIN; - /* Point to next chain block. */ - curr_sg = h->cmd_sg_list[c->cmdindex]->sgchain; + curr_sg = h->cmd_sg_list[c->cmdindex]; sg_index = 0; chained = 1; } @@ -3166,31 +3225,12 @@ static void do_cciss_request(struct request_queue *q) curr_sg[sg_index].Addr.lower = temp64.val32.lower; curr_sg[sg_index].Addr.upper = temp64.val32.upper; curr_sg[sg_index].Ext = 0; /* we are not chaining */ - ++sg_index; } - - if (chained) { - int len; - curr_sg = c->SG; - sg_index = h->max_cmd_sgentries - 1; - len = curr_sg[sg_index].Len; - /* Setup pointer to next chain block. - * Fill out last element in current chain - * block with address of next chain block. - */ - temp64.val = pci_map_single(h->pdev, - h->cmd_sg_list[c->cmdindex]->sgchain, - len, dir); - - h->cmd_sg_list[c->cmdindex]->sg_chain_dma = temp64.val; - curr_sg[sg_index].Addr.lower = temp64.val32.lower; - curr_sg[sg_index].Addr.upper = temp64.val32.upper; - - pci_dma_sync_single_for_device(h->pdev, - h->cmd_sg_list[c->cmdindex]->sg_chain_dma, - len, dir); - } + if (chained) + cciss_map_sg_chain_block(h, c, h->cmd_sg_list[c->cmdindex], + (seg - (h->max_cmd_sgentries - 1)) * + sizeof(SGDescriptor_struct)); /* track how many SG entries we are using */ if (seg > h->maxSG) @@ -3209,11 +3249,11 @@ static void do_cciss_request(struct request_queue *q) if (likely(blk_fs_request(creq))) { if(h->cciss_read == CCISS_READ_10) { c->Request.CDB[1] = 0; - c->Request.CDB[2] = (start_blk >> 24) & 0xff; //MSB + c->Request.CDB[2] = (start_blk >> 24) & 0xff; /* MSB */ c->Request.CDB[3] = (start_blk >> 16) & 0xff; c->Request.CDB[4] = (start_blk >> 8) & 0xff; c->Request.CDB[5] = start_blk & 0xff; - c->Request.CDB[6] = 0; // (sect >> 24) & 0xff; MSB + c->Request.CDB[6] = 0; /* (sect >> 24) & 0xff; MSB */ c->Request.CDB[7] = (blk_rq_sectors(creq) >> 8) & 0xff; c->Request.CDB[8] = blk_rq_sectors(creq) & 0xff; c->Request.CDB[9] = c->Request.CDB[11] = c->Request.CDB[12] = 0; @@ -3222,7 +3262,7 @@ static void do_cciss_request(struct request_queue *q) c->Request.CDBLen = 16; c->Request.CDB[1]= 0; - c->Request.CDB[2]= (upper32 >> 24) & 0xff; //MSB + c->Request.CDB[2]= (upper32 >> 24) & 0xff; /* MSB */ c->Request.CDB[3]= (upper32 >> 16) & 0xff; c->Request.CDB[4]= (upper32 >> 8) & 0xff; c->Request.CDB[5]= upper32 & 0xff; @@ -3300,6 +3340,7 @@ static irqreturn_t do_cciss_intr(int irq, void *dev_id) printk(KERN_WARNING "cciss: controller cciss%d failed, stopping.\n", h->ctlr); + spin_unlock_irqrestore(CCISS_LOCK(h->ctlr), flags); fail_all_cmds(h->ctlr); return IRQ_HANDLED; } @@ -4240,37 +4281,10 @@ static int __devinit cciss_init_one(struct pci_dev *pdev, goto clean4; } } - hba[i]->cmd_sg_list = kmalloc(sizeof(struct Cmd_sg_list *) * - hba[i]->nr_cmds, - GFP_KERNEL); - if (!hba[i]->cmd_sg_list) { - printk(KERN_ERR "cciss%d: Cannot get memory for " - "s/g chaining.\n", i); + hba[i]->cmd_sg_list = cciss_allocate_sg_chain_blocks(hba[i], + hba[i]->chainsize, hba[i]->nr_cmds); + if (!hba[i]->cmd_sg_list && hba[i]->chainsize > 0) goto clean4; - } - /* Build up chain blocks for each command */ - if (hba[i]->chainsize > 0) { - for (j = 0; j < hba[i]->nr_cmds; j++) { - hba[i]->cmd_sg_list[j] = - kmalloc(sizeof(struct Cmd_sg_list), - GFP_KERNEL); - if (!hba[i]->cmd_sg_list[j]) { - printk(KERN_ERR "cciss%d: Cannot get memory " - "for chain block.\n", i); - goto clean4; - } - /* Need a block of chainsized s/g elements. */ - hba[i]->cmd_sg_list[j]->sgchain = - kmalloc((hba[i]->chainsize * - sizeof(SGDescriptor_struct)), - GFP_KERNEL); - if (!hba[i]->cmd_sg_list[j]->sgchain) { - printk(KERN_ERR "cciss%d: Cannot get memory " - "for s/g chains\n", i); - goto clean4; - } - } - } spin_lock_init(&hba[i]->lock); @@ -4329,16 +4343,7 @@ clean4: for (k = 0; k < hba[i]->nr_cmds; k++) kfree(hba[i]->scatter_list[k]); kfree(hba[i]->scatter_list); - /* Only free up extra s/g lists if controller supports them */ - if (hba[i]->chainsize > 0) { - for (j = 0; j < hba[i]->nr_cmds; j++) { - if (hba[i]->cmd_sg_list[j]) { - kfree(hba[i]->cmd_sg_list[j]->sgchain); - kfree(hba[i]->cmd_sg_list[j]); - } - } - kfree(hba[i]->cmd_sg_list); - } + cciss_free_sg_chain_blocks(hba[i]->cmd_sg_list, hba[i]->nr_cmds); if (hba[i]->cmd_pool) pci_free_consistent(hba[i]->pdev, hba[i]->nr_cmds * sizeof(CommandList_struct), @@ -4456,16 +4461,7 @@ static void __devexit cciss_remove_one(struct pci_dev *pdev) for (j = 0; j < hba[i]->nr_cmds; j++) kfree(hba[i]->scatter_list[j]); kfree(hba[i]->scatter_list); - /* Only free up extra s/g lists if controller supports them */ - if (hba[i]->chainsize > 0) { - for (j = 0; j < hba[i]->nr_cmds; j++) { - if (hba[i]->cmd_sg_list[j]) { - kfree(hba[i]->cmd_sg_list[j]->sgchain); - kfree(hba[i]->cmd_sg_list[j]); - } - } - kfree(hba[i]->cmd_sg_list); - } + cciss_free_sg_chain_blocks(hba[i]->cmd_sg_list, hba[i]->nr_cmds); /* * Deliberately omit pci_disable_device(): it does something nasty to * Smart Array controllers that pci_enable_device does not undo @@ -4498,7 +4494,7 @@ static int __init cciss_init(void) * boundary. Given that we use pci_alloc_consistent() to allocate an * array of them, the size must be a multiple of 8 bytes. */ - BUILD_BUG_ON(sizeof(CommandList_struct) % 8); + BUILD_BUG_ON(sizeof(CommandList_struct) % COMMANDLIST_ALIGNMENT); printk(KERN_INFO DRIVER_NAME "\n"); diff --git a/drivers/block/cciss.h b/drivers/block/cciss.h index 1d95db254069..c5d411174db0 100644 --- a/drivers/block/cciss.h +++ b/drivers/block/cciss.h @@ -55,18 +55,12 @@ typedef struct _drive_info_struct char device_initialized; /* indicates whether dev is initialized */ } drive_info_struct; -struct Cmd_sg_list { - SGDescriptor_struct *sgchain; - dma_addr_t sg_chain_dma; - int chain_block_size; -}; - struct ctlr_info { int ctlr; char devname[8]; char *product_name; - char firm_ver[4]; // Firmware version + char firm_ver[4]; /* Firmware version */ struct pci_dev *pdev; __u32 board_id; void __iomem *vaddr; @@ -89,7 +83,7 @@ struct ctlr_info int maxsgentries; int chainsize; int max_cmd_sgentries; - struct Cmd_sg_list **cmd_sg_list; + SGDescriptor_struct **cmd_sg_list; # define DOORBELL_INT 0 # define PERF_MODE_INT 1 @@ -103,7 +97,7 @@ struct ctlr_info BYTE cciss_write; BYTE cciss_read_capacity; - // information about each logical volume + /* information about each logical volume */ drive_info_struct *drv[CISS_MAX_LUN]; struct access_method access; @@ -116,7 +110,7 @@ struct ctlr_info unsigned int maxSG; spinlock_t lock; - //* pointers to command and error info pool */ + /* pointers to command and error info pool */ CommandList_struct *cmd_pool; dma_addr_t cmd_pool_dhandle; ErrorInfo_struct *errinfo_pool; @@ -134,12 +128,10 @@ struct ctlr_info */ int next_to_run; - // Disk structures we need to pass back + /* Disk structures we need to pass back */ struct gendisk *gendisk[CISS_MAX_LUN]; #ifdef CONFIG_CISS_SCSI_TAPE - void *scsi_ctlr; /* ptr to structure containing scsi related stuff */ - /* list of block side commands the scsi error handling sucked up */ - /* and saved for later processing */ + struct cciss_scsi_adapter_data_t *scsi_ctlr; #endif unsigned char alive; struct list_head scan_list; @@ -315,4 +307,3 @@ struct board_type { #define CCISS_LOCK(i) (&hba[i]->lock) #endif /* CCISS_H */ - diff --git a/drivers/block/cciss_cmd.h b/drivers/block/cciss_cmd.h index 6afa700890ff..e624ff959cb6 100644 --- a/drivers/block/cciss_cmd.h +++ b/drivers/block/cciss_cmd.h @@ -1,31 +1,16 @@ #ifndef CCISS_CMD_H #define CCISS_CMD_H -//########################################################################### -//DEFINES -//########################################################################### + +#include <linux/cciss_defs.h> + +/* DEFINES */ #define CISS_VERSION "1.00" -//general boundary definitions -#define SENSEINFOBYTES 32//note that this value may vary between host implementations +/* general boundary definitions */ #define MAXSGENTRIES 32 #define CCISS_SG_CHAIN 0x80000000 #define MAXREPLYQS 256 -//Command Status value -#define CMD_SUCCESS 0x0000 -#define CMD_TARGET_STATUS 0x0001 -#define CMD_DATA_UNDERRUN 0x0002 -#define CMD_DATA_OVERRUN 0x0003 -#define CMD_INVALID 0x0004 -#define CMD_PROTOCOL_ERR 0x0005 -#define CMD_HARDWARE_ERR 0x0006 -#define CMD_CONNECTION_LOST 0x0007 -#define CMD_ABORTED 0x0008 -#define CMD_ABORT_FAILED 0x0009 -#define CMD_UNSOLICITED_ABORT 0x000A -#define CMD_TIMEOUT 0x000B -#define CMD_UNABORTABLE 0x000C - /* Unit Attentions ASC's as defined for the MSA2012sa */ #define POWER_OR_RESET 0x29 #define STATE_CHANGED 0x2a @@ -49,30 +34,13 @@ #define ASYM_ACCESS_CHANGED 0x06 #define LUN_CAPACITY_CHANGED 0x09 -//transfer direction -#define XFER_NONE 0x00 -#define XFER_WRITE 0x01 -#define XFER_READ 0x02 -#define XFER_RSVD 0x03 - -//task attribute -#define ATTR_UNTAGGED 0x00 -#define ATTR_SIMPLE 0x04 -#define ATTR_HEADOFQUEUE 0x05 -#define ATTR_ORDERED 0x06 -#define ATTR_ACA 0x07 - -//cdb type -#define TYPE_CMD 0x00 -#define TYPE_MSG 0x01 - -//config space register offsets +/* config space register offsets */ #define CFG_VENDORID 0x00 #define CFG_DEVICEID 0x02 #define CFG_I2OBAR 0x10 #define CFG_MEM1BAR 0x14 -//i2o space register offsets +/* i2o space register offsets */ #define I2O_IBDB_SET 0x20 #define I2O_IBDB_CLEAR 0x70 #define I2O_INT_STATUS 0x30 @@ -81,7 +49,7 @@ #define I2O_OBPOST_Q 0x44 #define I2O_DMA1_CFG 0x214 -//Configuration Table +/* Configuration Table */ #define CFGTBL_ChangeReq 0x00000001l #define CFGTBL_AccCmds 0x00000001l @@ -103,24 +71,17 @@ typedef union _u64bit __u64 val; } u64bit; -// Type defs used in the following structs -#define BYTE __u8 -#define WORD __u16 -#define HWORD __u16 -#define DWORD __u32 +/* Type defs used in the following structs */ #define QWORD vals32 -//########################################################################### -//STRUCTURES -//########################################################################### -#define CISS_MAX_LUN 1024 +/* STRUCTURES */ #define CISS_MAX_PHYS_LUN 1024 -// SCSI-3 Cmmands +/* SCSI-3 Cmmands */ #pragma pack(1) #define CISS_INQUIRY 0x12 -//Date returned +/* Date returned */ typedef struct _InquiryData_struct { BYTE data_byte[36]; @@ -128,7 +89,7 @@ typedef struct _InquiryData_struct #define CISS_REPORT_LOG 0xc2 /* Report Logical LUNs */ #define CISS_REPORT_PHYS 0xc3 /* Report Physical LUNs */ -// Data returned +/* Data returned */ typedef struct _ReportLUNdata_struct { BYTE LUNListLength[4]; @@ -139,8 +100,8 @@ typedef struct _ReportLUNdata_struct #define CCISS_READ_CAPACITY 0x25 /* Read Capacity */ typedef struct _ReadCapdata_struct { - BYTE total_size[4]; // Total size in blocks - BYTE block_size[4]; // Size of blocks in bytes + BYTE total_size[4]; /* Total size in blocks */ + BYTE block_size[4]; /* Size of blocks in bytes */ } ReadCapdata_struct; #define CCISS_READ_CAPACITY_16 0x9e /* Read Capacity 16 */ @@ -172,52 +133,13 @@ typedef struct _ReadCapdata_struct_16 #define CDB_LEN10 10 #define CDB_LEN16 16 -// BMIC commands +/* BMIC commands */ #define BMIC_READ 0x26 #define BMIC_WRITE 0x27 #define BMIC_CACHE_FLUSH 0xc2 -#define CCISS_CACHE_FLUSH 0x01 //C2 was already being used by CCISS - -//Command List Structure -typedef union _SCSI3Addr_struct { - struct { - BYTE Dev; - BYTE Bus:6; - BYTE Mode:2; // b00 - } PeripDev; - struct { - BYTE DevLSB; - BYTE DevMSB:6; - BYTE Mode:2; // b01 - } LogDev; - struct { - BYTE Dev:5; - BYTE Bus:3; - BYTE Targ:6; - BYTE Mode:2; // b10 - } LogUnit; -} SCSI3Addr_struct; - -typedef struct _PhysDevAddr_struct { - DWORD TargetId:24; - DWORD Bus:6; - DWORD Mode:2; - SCSI3Addr_struct Target[2]; //2 level target device addr -} PhysDevAddr_struct; - -typedef struct _LogDevAddr_struct { - DWORD VolId:30; - DWORD Mode:2; - BYTE reserved[4]; -} LogDevAddr_struct; - -typedef union _LUNAddr_struct { - BYTE LunAddrBytes[8]; - SCSI3Addr_struct SCSI3Lun[4]; - PhysDevAddr_struct PhysDev; - LogDevAddr_struct LogDev; -} LUNAddr_struct; +#define CCISS_CACHE_FLUSH 0x01 /* C2 was already being used by CCISS */ +/* Command List Structure */ #define CTLR_LUNID "\0\0\0\0\0\0\0\0" typedef struct _CommandListHeader_struct { @@ -227,16 +149,6 @@ typedef struct _CommandListHeader_struct { QWORD Tag; LUNAddr_struct LUN; } CommandListHeader_struct; -typedef struct _RequestBlock_struct { - BYTE CDBLen; - struct { - BYTE Type:3; - BYTE Attribute:3; - BYTE Direction:2; - } Type; - HWORD Timeout; - BYTE CDB[16]; -} RequestBlock_struct; typedef struct _ErrDescriptor_struct { QWORD Addr; DWORD Len; @@ -247,28 +159,6 @@ typedef struct _SGDescriptor_struct { DWORD Ext; } SGDescriptor_struct; -typedef union _MoreErrInfo_struct{ - struct { - BYTE Reserved[3]; - BYTE Type; - DWORD ErrorInfo; - }Common_Info; - struct{ - BYTE Reserved[2]; - BYTE offense_size;//size of offending entry - BYTE offense_num; //byte # of offense 0-base - DWORD offense_value; - }Invalid_Cmd; -}MoreErrInfo_struct; -typedef struct _ErrorInfo_struct { - BYTE ScsiStatus; - BYTE SenseLen; - HWORD CommandStatus; - DWORD ResidualCnt; - MoreErrInfo_struct MoreErrInfo; - BYTE SenseInfo[SENSEINFOBYTES]; -} ErrorInfo_struct; - /* Command types */ #define CMD_RWREQ 0x00 #define CMD_IOCTL_PEND 0x01 @@ -277,10 +167,18 @@ typedef struct _ErrorInfo_struct { #define CMD_MSG_TIMEOUT 0x05 #define CMD_MSG_STALE 0xff -/* This structure needs to be divisible by 8 for new - * indexing method. +/* This structure needs to be divisible by COMMANDLIST_ALIGNMENT + * because low bits of the address are used to to indicate that + * whether the tag contains an index or an address. PAD_32 and + * PAD_64 can be adjusted independently as needed for 32-bit + * and 64-bits systems. */ -#define PADSIZE (sizeof(long) - 4) +#define COMMANDLIST_ALIGNMENT (8) +#define IS_64_BIT ((sizeof(long) - 4)/4) +#define IS_32_BIT (!IS_64_BIT) +#define PAD_32 (0) +#define PAD_64 (4) +#define PADSIZE (IS_32_BIT * PAD_32 + IS_64_BIT * PAD_64) typedef struct _CommandList_struct { CommandListHeader_struct Header; RequestBlock_struct Request; @@ -300,7 +198,7 @@ typedef struct _CommandList_struct { char pad[PADSIZE]; } CommandList_struct; -//Configuration Table Structure +/* Configuration Table Structure */ typedef struct _HostWrite_struct { DWORD TransportRequest; DWORD Reserved; @@ -326,4 +224,4 @@ typedef struct _CfgTable_struct { DWORD MaxPhysicalDrivesPerLogicalUnit; } CfgTable_struct; #pragma pack() -#endif // CCISS_CMD_H +#endif /* CCISS_CMD_H */ diff --git a/drivers/block/cciss_scsi.c b/drivers/block/cciss_scsi.c index 5d0e46dc3632..72dae92f3cab 100644 --- a/drivers/block/cciss_scsi.c +++ b/drivers/block/cciss_scsi.c @@ -84,7 +84,6 @@ static struct scsi_host_template cciss_driver_template = { .queuecommand = cciss_scsi_queue_command, .can_queue = SCSI_CCISS_CAN_QUEUE, .this_id = 7, - .sg_tablesize = MAXSGENTRIES, .cmd_per_lun = 1, .use_clustering = DISABLE_CLUSTERING, /* Can't have eh_bus_reset_handler or eh_host_reset_handler for cciss */ @@ -93,11 +92,16 @@ static struct scsi_host_template cciss_driver_template = { }; #pragma pack(1) + +#define SCSI_PAD_32 0 +#define SCSI_PAD_64 0 + struct cciss_scsi_cmd_stack_elem_t { CommandList_struct cmd; ErrorInfo_struct Err; __u32 busaddr; - __u32 pad; + int cmdindex; + u8 pad[IS_32_BIT * SCSI_PAD_32 + IS_64_BIT * SCSI_PAD_64]; }; #pragma pack() @@ -118,16 +122,15 @@ struct cciss_scsi_cmd_stack_t { struct cciss_scsi_adapter_data_t { struct Scsi_Host *scsi_host; struct cciss_scsi_cmd_stack_t cmd_stack; + SGDescriptor_struct **cmd_sg_list; int registered; spinlock_t lock; // to protect ccissscsi[ctlr]; }; #define CPQ_TAPE_LOCK(ctlr, flags) spin_lock_irqsave( \ - &(((struct cciss_scsi_adapter_data_t *) \ - hba[ctlr]->scsi_ctlr)->lock), flags); + &hba[ctlr]->scsi_ctlr->lock, flags); #define CPQ_TAPE_UNLOCK(ctlr, flags) spin_unlock_irqrestore( \ - &(((struct cciss_scsi_adapter_data_t *) \ - hba[ctlr]->scsi_ctlr)->lock), flags); + &hba[ctlr]->scsi_ctlr->lock, flags); static CommandList_struct * scsi_cmd_alloc(ctlr_info_t *h) @@ -143,7 +146,7 @@ scsi_cmd_alloc(ctlr_info_t *h) struct cciss_scsi_cmd_stack_t *stk; u64bit temp64; - sa = (struct cciss_scsi_adapter_data_t *) h->scsi_ctlr; + sa = h->scsi_ctlr; stk = &sa->cmd_stack; if (stk->top < 0) @@ -154,6 +157,7 @@ scsi_cmd_alloc(ctlr_info_t *h) memset(&c->Err, 0, sizeof(c->Err)); /* set physical addr of cmd and addr of scsi parameters */ c->cmd.busaddr = c->busaddr; + c->cmd.cmdindex = c->cmdindex; /* (__u32) (stk->cmd_pool_handle + (sizeof(struct cciss_scsi_cmd_stack_elem_t)*stk->top)); */ @@ -182,13 +186,13 @@ scsi_cmd_free(ctlr_info_t *h, CommandList_struct *cmd) struct cciss_scsi_adapter_data_t *sa; struct cciss_scsi_cmd_stack_t *stk; - sa = (struct cciss_scsi_adapter_data_t *) h->scsi_ctlr; + sa = h->scsi_ctlr; stk = &sa->cmd_stack; + stk->top++; if (stk->top >= CMD_STACK_SIZE) { printk("cciss: scsi_cmd_free called too many times.\n"); BUG(); } - stk->top++; stk->elem[stk->top] = (struct cciss_scsi_cmd_stack_elem_t *) cmd; } @@ -199,24 +203,31 @@ scsi_cmd_stack_setup(int ctlr, struct cciss_scsi_adapter_data_t *sa) struct cciss_scsi_cmd_stack_t *stk; size_t size; + sa->cmd_sg_list = cciss_allocate_sg_chain_blocks(hba[ctlr], + hba[ctlr]->chainsize, CMD_STACK_SIZE); + if (!sa->cmd_sg_list && hba[ctlr]->chainsize > 0) + return -ENOMEM; + stk = &sa->cmd_stack; size = sizeof(struct cciss_scsi_cmd_stack_elem_t) * CMD_STACK_SIZE; - // pci_alloc_consistent guarantees 32-bit DMA address will - // be used - + /* Check alignment, see cciss_cmd.h near CommandList_struct def. */ + BUILD_BUG_ON((sizeof(*stk->pool) % COMMANDLIST_ALIGNMENT) != 0); + /* pci_alloc_consistent guarantees 32-bit DMA address will be used */ stk->pool = (struct cciss_scsi_cmd_stack_elem_t *) pci_alloc_consistent(hba[ctlr]->pdev, size, &stk->cmd_pool_handle); if (stk->pool == NULL) { - printk("stk->pool is null\n"); - return -1; + cciss_free_sg_chain_blocks(sa->cmd_sg_list, CMD_STACK_SIZE); + sa->cmd_sg_list = NULL; + return -ENOMEM; } for (i=0; i<CMD_STACK_SIZE; i++) { stk->elem[i] = &stk->pool[i]; stk->elem[i]->busaddr = (__u32) (stk->cmd_pool_handle + (sizeof(struct cciss_scsi_cmd_stack_elem_t) * i)); + stk->elem[i]->cmdindex = i; } stk->top = CMD_STACK_SIZE-1; return 0; @@ -229,7 +240,7 @@ scsi_cmd_stack_free(int ctlr) struct cciss_scsi_cmd_stack_t *stk; size_t size; - sa = (struct cciss_scsi_adapter_data_t *) hba[ctlr]->scsi_ctlr; + sa = hba[ctlr]->scsi_ctlr; stk = &sa->cmd_stack; if (stk->top != CMD_STACK_SIZE-1) { printk( "cciss: %d scsi commands are still outstanding.\n", @@ -241,6 +252,7 @@ scsi_cmd_stack_free(int ctlr) pci_free_consistent(hba[ctlr]->pdev, size, stk->pool, stk->cmd_pool_handle); stk->pool = NULL; + cciss_free_sg_chain_blocks(sa->cmd_sg_list, CMD_STACK_SIZE); } #if 0 @@ -530,8 +542,7 @@ adjust_cciss_scsi_table(int ctlr, int hostno, CPQ_TAPE_LOCK(ctlr, flags); if (hostno != -1) /* if it's not the first time... */ - sh = ((struct cciss_scsi_adapter_data_t *) - hba[ctlr]->scsi_ctlr)->scsi_host; + sh = hba[ctlr]->scsi_ctlr->scsi_host; /* find any devices in ccissscsi[] that are not in sd[] and remove them from ccissscsi[] */ @@ -702,7 +713,7 @@ cciss_scsi_setup(int cntl_num) kfree(shba); shba = NULL; } - hba[cntl_num]->scsi_ctlr = (void *) shba; + hba[cntl_num]->scsi_ctlr = shba; return; } @@ -725,6 +736,8 @@ complete_scsi_command( CommandList_struct *cp, int timeout, __u32 tag) ctlr = hba[cp->ctlr]; scsi_dma_unmap(cmd); + if (cp->Header.SGTotal > ctlr->max_cmd_sgentries) + cciss_unmap_sg_chain_block(ctlr, cp); cmd->result = (DID_OK << 16); /* host byte */ cmd->result |= (COMMAND_COMPLETE << 8); /* msg byte */ @@ -847,9 +860,11 @@ cciss_scsi_detect(int ctlr) sh->io_port = 0; // good enough? FIXME, sh->n_io_port = 0; // I don't think we use these two... sh->this_id = SELF_SCSI_ID; + sh->sg_tablesize = hba[ctlr]->maxsgentries; + sh->max_cmd_len = MAX_COMMAND_SIZE; ((struct cciss_scsi_adapter_data_t *) - hba[ctlr]->scsi_ctlr)->scsi_host = (void *) sh; + hba[ctlr]->scsi_ctlr)->scsi_host = sh; sh->hostdata[0] = (unsigned long) hba[ctlr]; sh->irq = hba[ctlr]->intr[SIMPLE_MODE_INT]; sh->unique_id = sh->irq; @@ -1364,34 +1379,54 @@ cciss_scsi_proc_info(struct Scsi_Host *sh, dma mapping and fills in the scatter gather entries of the cciss command, cp. */ -static void -cciss_scatter_gather(struct pci_dev *pdev, - CommandList_struct *cp, - struct scsi_cmnd *cmd) +static void cciss_scatter_gather(ctlr_info_t *h, CommandList_struct *cp, + struct scsi_cmnd *cmd) { unsigned int len; struct scatterlist *sg; __u64 addr64; - int use_sg, i; - - BUG_ON(scsi_sg_count(cmd) > MAXSGENTRIES); - - use_sg = scsi_dma_map(cmd); - if (use_sg) { /* not too many addrs? */ - scsi_for_each_sg(cmd, sg, use_sg, i) { + int request_nsgs, i, chained, sg_index; + struct cciss_scsi_adapter_data_t *sa = h->scsi_ctlr; + SGDescriptor_struct *curr_sg; + + BUG_ON(scsi_sg_count(cmd) > h->maxsgentries); + + chained = 0; + sg_index = 0; + curr_sg = cp->SG; + request_nsgs = scsi_dma_map(cmd); + if (request_nsgs) { + scsi_for_each_sg(cmd, sg, request_nsgs, i) { + if (sg_index + 1 == h->max_cmd_sgentries && + !chained && request_nsgs - i > 1) { + chained = 1; + sg_index = 0; + curr_sg = sa->cmd_sg_list[cp->cmdindex]; + } addr64 = (__u64) sg_dma_address(sg); len = sg_dma_len(sg); - cp->SG[i].Addr.lower = - (__u32) (addr64 & (__u64) 0x00000000FFFFFFFF); - cp->SG[i].Addr.upper = - (__u32) ((addr64 >> 32) & (__u64) 0x00000000FFFFFFFF); - cp->SG[i].Len = len; - cp->SG[i].Ext = 0; // we are not chaining + curr_sg[sg_index].Addr.lower = + (__u32) (addr64 & 0x0FFFFFFFFULL); + curr_sg[sg_index].Addr.upper = + (__u32) ((addr64 >> 32) & 0x0FFFFFFFFULL); + curr_sg[sg_index].Len = len; + curr_sg[sg_index].Ext = 0; + ++sg_index; } + if (chained) + cciss_map_sg_chain_block(h, cp, + sa->cmd_sg_list[cp->cmdindex], + (request_nsgs - (h->max_cmd_sgentries - 1)) * + sizeof(SGDescriptor_struct)); } - - cp->Header.SGList = (__u8) use_sg; /* no. SGs contig in this cmd */ - cp->Header.SGTotal = (__u16) use_sg; /* total sgs in this cmd list */ + /* track how many SG entries we are using */ + if (request_nsgs > h->maxSG) + h->maxSG = request_nsgs; + cp->Header.SGTotal = (__u8) request_nsgs + chained; + if (request_nsgs > h->max_cmd_sgentries) + cp->Header.SGList = h->max_cmd_sgentries; + else + cp->Header.SGList = cp->Header.SGTotal; return; } @@ -1399,7 +1434,7 @@ cciss_scatter_gather(struct pci_dev *pdev, static int cciss_scsi_queue_command (struct scsi_cmnd *cmd, void (* done)(struct scsi_cmnd *)) { - ctlr_info_t **c; + ctlr_info_t *c; int ctlr, rc; unsigned char scsi3addr[8]; CommandList_struct *cp; @@ -1407,8 +1442,8 @@ cciss_scsi_queue_command (struct scsi_cmnd *cmd, void (* done)(struct scsi_cmnd // Get the ptr to our adapter structure (hba[i]) out of cmd->host. // We violate cmd->host privacy here. (Is there another way?) - c = (ctlr_info_t **) &cmd->device->host->hostdata[0]; - ctlr = (*c)->ctlr; + c = (ctlr_info_t *) cmd->device->host->hostdata[0]; + ctlr = c->ctlr; rc = lookup_scsi3addr(ctlr, cmd->device->channel, cmd->device->id, cmd->device->lun, scsi3addr); @@ -1431,7 +1466,7 @@ cciss_scsi_queue_command (struct scsi_cmnd *cmd, void (* done)(struct scsi_cmnd see what the device thinks of it. */ spin_lock_irqsave(CCISS_LOCK(ctlr), flags); - cp = scsi_cmd_alloc(*c); + cp = scsi_cmd_alloc(c); spin_unlock_irqrestore(CCISS_LOCK(ctlr), flags); if (cp == NULL) { /* trouble... */ printk("scsi_cmd_alloc returned NULL!\n"); @@ -1489,15 +1524,14 @@ cciss_scsi_queue_command (struct scsi_cmnd *cmd, void (* done)(struct scsi_cmnd BUG(); break; } - - cciss_scatter_gather((*c)->pdev, cp, cmd); // Fill the SG list + cciss_scatter_gather(c, cp, cmd); /* Put the request on the tail of the request queue */ spin_lock_irqsave(CCISS_LOCK(ctlr), flags); - addQ(&(*c)->reqQ, cp); - (*c)->Qdepth++; - start_io(*c); + addQ(&c->reqQ, cp); + c->Qdepth++; + start_io(c); spin_unlock_irqrestore(CCISS_LOCK(ctlr), flags); /* the cmd'll come back via intr handler in complete_scsi_command() */ @@ -1514,7 +1548,7 @@ cciss_unregister_scsi(int ctlr) /* we are being forcibly unloaded, and may not refuse. */ spin_lock_irqsave(CCISS_LOCK(ctlr), flags); - sa = (struct cciss_scsi_adapter_data_t *) hba[ctlr]->scsi_ctlr; + sa = hba[ctlr]->scsi_ctlr; stk = &sa->cmd_stack; /* if we weren't ever actually registered, don't unregister */ @@ -1541,7 +1575,7 @@ cciss_engage_scsi(int ctlr) unsigned long flags; spin_lock_irqsave(CCISS_LOCK(ctlr), flags); - sa = (struct cciss_scsi_adapter_data_t *) hba[ctlr]->scsi_ctlr; + sa = hba[ctlr]->scsi_ctlr; stk = &sa->cmd_stack; if (sa->registered) { @@ -1654,14 +1688,14 @@ static int cciss_eh_device_reset_handler(struct scsi_cmnd *scsicmd) int rc; CommandList_struct *cmd_in_trouble; unsigned char lunaddr[8]; - ctlr_info_t **c; + ctlr_info_t *c; int ctlr; /* find the controller to which the command to be aborted was sent */ - c = (ctlr_info_t **) &scsicmd->device->host->hostdata[0]; + c = (ctlr_info_t *) scsicmd->device->host->hostdata[0]; if (c == NULL) /* paranoia */ return FAILED; - ctlr = (*c)->ctlr; + ctlr = c->ctlr; printk(KERN_WARNING "cciss%d: resetting tape drive or medium changer.\n", ctlr); /* find the command that's giving us trouble */ cmd_in_trouble = (CommandList_struct *) scsicmd->host_scribble; @@ -1671,7 +1705,7 @@ static int cciss_eh_device_reset_handler(struct scsi_cmnd *scsicmd) /* send a reset to the SCSI LUN which the command was sent to */ rc = sendcmd_withirq(CCISS_RESET_MSG, ctlr, NULL, 0, 0, lunaddr, TYPE_MSG); - if (rc == 0 && wait_for_device_to_become_ready(*c, lunaddr) == 0) + if (rc == 0 && wait_for_device_to_become_ready(c, lunaddr) == 0) return SUCCESS; printk(KERN_WARNING "cciss%d: resetting device failed.\n", ctlr); return FAILED; @@ -1682,14 +1716,14 @@ static int cciss_eh_abort_handler(struct scsi_cmnd *scsicmd) int rc; CommandList_struct *cmd_to_abort; unsigned char lunaddr[8]; - ctlr_info_t **c; + ctlr_info_t *c; int ctlr; /* find the controller to which the command to be aborted was sent */ - c = (ctlr_info_t **) &scsicmd->device->host->hostdata[0]; + c = (ctlr_info_t *) scsicmd->device->host->hostdata[0]; if (c == NULL) /* paranoia */ return FAILED; - ctlr = (*c)->ctlr; + ctlr = c->ctlr; printk(KERN_WARNING "cciss%d: aborting tardy SCSI cmd\n", ctlr); /* find the command to be aborted */ diff --git a/drivers/block/cciss_scsi.h b/drivers/block/cciss_scsi.h index 7b750245ae76..6d5822fe851a 100644 --- a/drivers/block/cciss_scsi.h +++ b/drivers/block/cciss_scsi.h @@ -25,16 +25,16 @@ #include <scsi/scsicam.h> /* possibly irrelevant, since we don't show disks */ - // the scsi id of the adapter... + /* the scsi id of the adapter... */ #define SELF_SCSI_ID 15 - // 15 is somewhat arbitrary, since the scsi-2 bus - // that's presented by the driver to the OS is - // fabricated. The "real" scsi-3 bus the - // hardware presents is fabricated too. - // The actual, honest-to-goodness physical - // bus that the devices are attached to is not - // addressible natively, and may in fact turn - // out to be not scsi at all. + /* 15 is somewhat arbitrary, since the scsi-2 bus + that's presented by the driver to the OS is + fabricated. The "real" scsi-3 bus the + hardware presents is fabricated too. + The actual, honest-to-goodness physical + bus that the devices are attached to is not + addressible natively, and may in fact turn + out to be not scsi at all. */ #define SCSI_CCISS_CAN_QUEUE 2 diff --git a/drivers/block/cpqarray.c b/drivers/block/cpqarray.c index 6422651ec364..abb4ec6690fc 100644 --- a/drivers/block/cpqarray.c +++ b/drivers/block/cpqarray.c @@ -386,7 +386,7 @@ static void __devexit cpqarray_remove_one_eisa (int i) } /* pdev is NULL for eisa */ -static int __init cpqarray_register_ctlr( int i, struct pci_dev *pdev) +static int __devinit cpqarray_register_ctlr( int i, struct pci_dev *pdev) { struct request_queue *q; int j; @@ -448,11 +448,8 @@ static int __init cpqarray_register_ctlr( int i, struct pci_dev *pdev) blk_queue_bounce_limit(q, hba[i]->pci_dev->dma_mask); /* This is a hardware imposed limit. */ - blk_queue_max_hw_segments(q, SG_MAX); + blk_queue_max_segments(q, SG_MAX); - /* This is a driver limit and could be eliminated. */ - blk_queue_max_phys_segments(q, SG_MAX); - init_timer(&hba[i]->timer); hba[i]->timer.expires = jiffies + IDA_TIMER; hba[i]->timer.data = (unsigned long)hba[i]; @@ -506,7 +503,7 @@ Enomem4: return -1; } -static int __init cpqarray_init_one( struct pci_dev *pdev, +static int __devinit cpqarray_init_one( struct pci_dev *pdev, const struct pci_device_id *ent) { int i; @@ -743,7 +740,7 @@ __setup("smart2=", cpqarray_setup); /* * Find an EISA controller's signature. Set up an hba if we find it. */ -static int __init cpqarray_eisa_detect(void) +static int __devinit cpqarray_eisa_detect(void) { int i=0, j; __u32 board_id; diff --git a/drivers/block/drbd/drbd_actlog.c b/drivers/block/drbd/drbd_actlog.c index 17956ff6a08d..df018990c422 100644 --- a/drivers/block/drbd/drbd_actlog.c +++ b/drivers/block/drbd/drbd_actlog.c @@ -536,7 +536,9 @@ static void atodb_endio(struct bio *bio, int error) put_ldev(mdev); } +/* sector to word */ #define S2W(s) ((s)<<(BM_EXT_SHIFT-BM_BLOCK_SHIFT-LN2_BPL)) + /* activity log to on disk bitmap -- prepare bio unless that sector * is already covered by previously prepared bios */ static int atodb_prepare_unless_covered(struct drbd_conf *mdev, @@ -546,13 +548,20 @@ static int atodb_prepare_unless_covered(struct drbd_conf *mdev, { struct bio *bio; struct page *page; - sector_t on_disk_sector = enr + mdev->ldev->md.md_offset - + mdev->ldev->md.bm_offset; + sector_t on_disk_sector; unsigned int page_offset = PAGE_SIZE; int offset; int i = 0; int err = -ENOMEM; + /* We always write aligned, full 4k blocks, + * so we can ignore the logical_block_size (for now) */ + enr &= ~7U; + on_disk_sector = enr + mdev->ldev->md.md_offset + + mdev->ldev->md.bm_offset; + + D_ASSERT(!(on_disk_sector & 7U)); + /* Check if that enr is already covered by an already created bio. * Caution, bios[] is not NULL terminated, * but only initialized to all NULL. @@ -588,7 +597,7 @@ static int atodb_prepare_unless_covered(struct drbd_conf *mdev, offset = S2W(enr); drbd_bm_get_lel(mdev, offset, - min_t(size_t, S2W(1), drbd_bm_words(mdev) - offset), + min_t(size_t, S2W(8), drbd_bm_words(mdev) - offset), kmap(page) + page_offset); kunmap(page); @@ -597,7 +606,7 @@ static int atodb_prepare_unless_covered(struct drbd_conf *mdev, bio->bi_bdev = mdev->ldev->md_bdev; bio->bi_sector = on_disk_sector; - if (bio_add_page(bio, page, MD_SECTOR_SIZE, page_offset) != MD_SECTOR_SIZE) + if (bio_add_page(bio, page, 4096, page_offset) != 4096) goto out_put_page; atomic_inc(&wc->count); @@ -1327,7 +1336,7 @@ int drbd_rs_del_all(struct drbd_conf *mdev) /* ok, ->resync is there. */ for (i = 0; i < mdev->resync->nr_elements; i++) { e = lc_element_by_index(mdev->resync, i); - bm_ext = e ? lc_entry(e, struct bm_extent, lce) : NULL; + bm_ext = lc_entry(e, struct bm_extent, lce); if (bm_ext->lce.lc_number == LC_FREE) continue; if (bm_ext->lce.lc_number == mdev->resync_wenr) { diff --git a/drivers/block/drbd/drbd_bitmap.c b/drivers/block/drbd/drbd_bitmap.c index b61057e77882..e3f88d6e1412 100644 --- a/drivers/block/drbd/drbd_bitmap.c +++ b/drivers/block/drbd/drbd_bitmap.c @@ -26,6 +26,7 @@ #include <linux/vmalloc.h> #include <linux/string.h> #include <linux/drbd.h> +#include <linux/slab.h> #include <asm/kmap_types.h> #include "drbd_int.h" @@ -66,7 +67,7 @@ struct drbd_bitmap { size_t bm_words; size_t bm_number_of_pages; sector_t bm_dev_capacity; - struct semaphore bm_change; /* serializes resize operations */ + struct mutex bm_change; /* serializes resize operations */ atomic_t bm_async_io; wait_queue_head_t bm_io_wait; @@ -83,6 +84,9 @@ struct drbd_bitmap { #define BM_MD_IO_ERROR 1 #define BM_P_VMALLOCED 2 +static int __bm_change_bits_to(struct drbd_conf *mdev, const unsigned long s, + unsigned long e, int val, const enum km_type km); + static int bm_is_locked(struct drbd_bitmap *b) { return test_bit(BM_LOCKED, &b->bm_flags); @@ -114,7 +118,7 @@ void drbd_bm_lock(struct drbd_conf *mdev, char *why) return; } - trylock_failed = down_trylock(&b->bm_change); + trylock_failed = !mutex_trylock(&b->bm_change); if (trylock_failed) { dev_warn(DEV, "%s going to '%s' but bitmap already locked for '%s' by %s\n", @@ -125,7 +129,7 @@ void drbd_bm_lock(struct drbd_conf *mdev, char *why) b->bm_task == mdev->receiver.task ? "receiver" : b->bm_task == mdev->asender.task ? "asender" : b->bm_task == mdev->worker.task ? "worker" : "?"); - down(&b->bm_change); + mutex_lock(&b->bm_change); } if (__test_and_set_bit(BM_LOCKED, &b->bm_flags)) dev_err(DEV, "FIXME bitmap already locked in bm_lock\n"); @@ -147,7 +151,7 @@ void drbd_bm_unlock(struct drbd_conf *mdev) b->bm_why = NULL; b->bm_task = NULL; - up(&b->bm_change); + mutex_unlock(&b->bm_change); } /* word offset to long pointer */ @@ -295,7 +299,7 @@ int drbd_bm_init(struct drbd_conf *mdev) if (!b) return -ENOMEM; spin_lock_init(&b->bm_lock); - init_MUTEX(&b->bm_change); + mutex_init(&b->bm_change); init_waitqueue_head(&b->bm_io_wait); mdev->bitmap = b; @@ -440,7 +444,7 @@ static void bm_memset(struct drbd_bitmap *b, size_t offset, int c, size_t len) * In case this is actually a resize, we copy the old bitmap into the new one. * Otherwise, the bitmap is initialized to all bits set. */ -int drbd_bm_resize(struct drbd_conf *mdev, sector_t capacity) +int drbd_bm_resize(struct drbd_conf *mdev, sector_t capacity, int set_new_bits) { struct drbd_bitmap *b = mdev->bitmap; unsigned long bits, words, owords, obits, *p_addr, *bm; @@ -515,7 +519,7 @@ int drbd_bm_resize(struct drbd_conf *mdev, sector_t capacity) obits = b->bm_bits; growing = bits > obits; - if (opages) + if (opages && growing && set_new_bits) bm_set_surplus(b); b->bm_pages = npages; @@ -525,8 +529,12 @@ int drbd_bm_resize(struct drbd_conf *mdev, sector_t capacity) b->bm_dev_capacity = capacity; if (growing) { - bm_memset(b, owords, 0xff, words-owords); - b->bm_set += bits - obits; + if (set_new_bits) { + bm_memset(b, owords, 0xff, words-owords); + b->bm_set += bits - obits; + } else + bm_memset(b, owords, 0x00, words-owords); + } if (want < have) { @@ -772,7 +780,7 @@ static void bm_page_io_async(struct drbd_conf *mdev, struct drbd_bitmap *b, int /* nothing to do, on disk == in memory */ # define bm_cpu_to_lel(x) ((void)0) # else -void bm_cpu_to_lel(struct drbd_bitmap *b) +static void bm_cpu_to_lel(struct drbd_bitmap *b) { /* need to cpu_to_lel all the pages ... * this may be optimized by using @@ -1014,7 +1022,7 @@ unsigned long _drbd_bm_find_next_zero(struct drbd_conf *mdev, unsigned long bm_f * wants bitnr, not sector. * expected to be called for only a few bits (e - s about BITS_PER_LONG). * Must hold bitmap lock already. */ -int __bm_change_bits_to(struct drbd_conf *mdev, const unsigned long s, +static int __bm_change_bits_to(struct drbd_conf *mdev, const unsigned long s, unsigned long e, int val, const enum km_type km) { struct drbd_bitmap *b = mdev->bitmap; @@ -1052,7 +1060,7 @@ int __bm_change_bits_to(struct drbd_conf *mdev, const unsigned long s, * for val != 0, we change 0 -> 1, return code positive * for val == 0, we change 1 -> 0, return code negative * wants bitnr, not sector */ -int bm_change_bits_to(struct drbd_conf *mdev, const unsigned long s, +static int bm_change_bits_to(struct drbd_conf *mdev, const unsigned long s, const unsigned long e, int val) { unsigned long flags; diff --git a/drivers/block/drbd/drbd_int.h b/drivers/block/drbd/drbd_int.h index 2bf3a6ef3684..485ed8c7d623 100644 --- a/drivers/block/drbd/drbd_int.h +++ b/drivers/block/drbd/drbd_int.h @@ -95,7 +95,7 @@ extern char usermode_helper[]; /* All EEs on the free list should have ID_VACANT (== 0) * freshly allocated EEs get !ID_VACANT (== 1) - * so if it says "cannot dereference null pointer at adress 0x00000001", + * so if it says "cannot dereference null pointer at address 0x00000001", * it is most likely one of these :( */ #define ID_IN_SYNC (4711ULL) @@ -132,6 +132,7 @@ enum { DRBD_FAULT_DT_RA = 6, /* data read ahead */ DRBD_FAULT_BM_ALLOC = 7, /* bitmap allocation */ DRBD_FAULT_AL_EE = 8, /* alloc ee */ + DRBD_FAULT_RECEIVE = 9, /* Changes some bytes upon receiving a [rs]data block */ DRBD_FAULT_MAX, }; @@ -208,8 +209,11 @@ enum drbd_packets { P_RS_IS_IN_SYNC = 0x22, /* meta socket */ P_SYNC_PARAM89 = 0x23, /* data socket, protocol version 89 replacement for P_SYNC_PARAM */ P_COMPRESSED_BITMAP = 0x24, /* compressed or otherwise encoded bitmap transfer */ + /* P_CKPT_FENCE_REQ = 0x25, * currently reserved for protocol D */ + /* P_CKPT_DISABLE_REQ = 0x26, * currently reserved for protocol D */ + P_DELAY_PROBE = 0x27, /* is used on BOTH sockets */ - P_MAX_CMD = 0x25, + P_MAX_CMD = 0x28, P_MAY_IGNORE = 0x100, /* Flag to test if (cmd > P_MAY_IGNORE) ... */ P_MAX_OPT_CMD = 0x101, @@ -261,6 +265,10 @@ static inline const char *cmdname(enum drbd_packets cmd) [P_OV_REQUEST] = "OVRequest", [P_OV_REPLY] = "OVReply", [P_OV_RESULT] = "OVResult", + [P_CSUM_RS_REQUEST] = "CsumRSRequest", + [P_RS_IS_IN_SYNC] = "CsumRSIsInSync", + [P_COMPRESSED_BITMAP] = "CBitmap", + [P_DELAY_PROBE] = "DelayProbe", [P_MAX_CMD] = NULL, }; @@ -443,13 +451,18 @@ struct p_rs_param_89 { char csums_alg[SHARED_SECRET_MAX]; } __packed; +enum drbd_conn_flags { + CF_WANT_LOSE = 1, + CF_DRY_RUN = 2, +}; + struct p_protocol { struct p_header head; u32 protocol; u32 after_sb_0p; u32 after_sb_1p; u32 after_sb_2p; - u32 want_lose; + u32 conn_flags; u32 two_primaries; /* Since protocol version 87 and higher. */ @@ -473,7 +486,8 @@ struct p_sizes { u64 u_size; /* user requested size */ u64 c_size; /* current exported size */ u32 max_segment_size; /* Maximal size of a BIO */ - u32 queue_order_type; + u16 queue_order_type; /* not yet implemented in DRBD*/ + u16 dds_flags; /* use enum dds_flags here. */ } __packed; struct p_state { @@ -530,6 +544,18 @@ struct p_compressed_bm { u8 code[0]; } __packed; +struct p_delay_probe { + struct p_header head; + u32 seq_num; /* sequence number to match the two probe packets */ + u32 offset; /* usecs the probe got sent after the reference time point */ +} __packed; + +struct delay_probe { + struct list_head list; + unsigned int seq_num; + struct timeval time; +}; + /* DCBP: Drbd Compressed Bitmap Packet ... */ static inline enum drbd_bitmap_code DCBP_get_code(struct p_compressed_bm *p) @@ -714,22 +740,6 @@ enum epoch_event { EV_CLEANUP = 32, /* used as flag */ }; -struct drbd_epoch_entry { - struct drbd_work w; - struct drbd_conf *mdev; - struct bio *private_bio; - struct hlist_node colision; - sector_t sector; - unsigned int size; - struct drbd_epoch *epoch; - - /* up to here, the struct layout is identical to drbd_request; - * we might be able to use that to our advantage... */ - - unsigned int flags; - u64 block_id; -}; - struct drbd_wq_barrier { struct drbd_work w; struct completion done; @@ -740,17 +750,49 @@ struct digest_info { void *digest; }; -/* ee flag bits */ +struct drbd_epoch_entry { + struct drbd_work w; + struct hlist_node colision; + struct drbd_epoch *epoch; + struct drbd_conf *mdev; + struct page *pages; + atomic_t pending_bios; + unsigned int size; + /* see comments on ee flag bits below */ + unsigned long flags; + sector_t sector; + u64 block_id; +}; + +/* ee flag bits. + * While corresponding bios are in flight, the only modification will be + * set_bit WAS_ERROR, which has to be atomic. + * If no bios are in flight yet, or all have been completed, + * non-atomic modification to ee->flags is ok. + */ enum { __EE_CALL_AL_COMPLETE_IO, - __EE_CONFLICT_PENDING, __EE_MAY_SET_IN_SYNC, + + /* This epoch entry closes an epoch using a barrier. + * On sucessful completion, the epoch is released, + * and the P_BARRIER_ACK send. */ __EE_IS_BARRIER, + + /* In case a barrier failed, + * we need to resubmit without the barrier flag. */ + __EE_RESUBMITTED, + + /* we may have several bios per epoch entry. + * if any of those fail, we set this flag atomically + * from the endio callback */ + __EE_WAS_ERROR, }; #define EE_CALL_AL_COMPLETE_IO (1<<__EE_CALL_AL_COMPLETE_IO) -#define EE_CONFLICT_PENDING (1<<__EE_CONFLICT_PENDING) #define EE_MAY_SET_IN_SYNC (1<<__EE_MAY_SET_IN_SYNC) #define EE_IS_BARRIER (1<<__EE_IS_BARRIER) +#define EE_RESUBMITTED (1<<__EE_RESUBMITTED) +#define EE_WAS_ERROR (1<<__EE_WAS_ERROR) /* global flag bits */ enum { @@ -791,6 +833,8 @@ enum { * while this is set. */ RESIZE_PENDING, /* Size change detected locally, waiting for the response from * the peer, if it changed there as well. */ + CONN_DRY_RUN, /* Expect disconnect after resync handshake. */ + GOT_PING_ACK, /* set when we receive a ping_ack packet, misc wait gets woken */ }; struct drbd_bitmap; /* opaque for drbd_conf */ @@ -898,9 +942,11 @@ struct drbd_conf { unsigned int ko_count; struct drbd_work resync_work, unplug_work, - md_sync_work; + md_sync_work, + delay_probe_work; struct timer_list resync_timer; struct timer_list md_sync_timer; + struct timer_list delay_probe_timer; /* Used after attach while negotiating new disk state. */ union drbd_state new_state_tmp; @@ -1016,6 +1062,12 @@ struct drbd_conf { u64 ed_uuid; /* UUID of the exposed data */ struct mutex state_mutex; char congestion_reason; /* Why we where congested... */ + struct list_head delay_probes; /* protected by peer_seq_lock */ + int data_delay; /* Delay of packets on the data-sock behind meta-sock */ + unsigned int delay_seq; /* To generate sequence numbers of delay probes */ + struct timeval dps_time; /* delay-probes-start-time */ + unsigned int dp_volume_last; /* send_cnt of last delay probe */ + int c_sync_rate; /* current resync rate after delay_probe magic */ }; static inline struct drbd_conf *minor_to_mdev(unsigned int minor) @@ -1071,6 +1123,11 @@ enum chg_state_flags { CS_ORDERED = CS_WAIT_COMPLETE + CS_SERIALIZE, }; +enum dds_flags { + DDSF_FORCED = 1, + DDSF_NO_RESYNC = 2, /* Do not run a resync for the new space */ +}; + extern void drbd_init_set_defaults(struct drbd_conf *mdev); extern int drbd_change_state(struct drbd_conf *mdev, enum chg_state_flags f, union drbd_state mask, union drbd_state val); @@ -1103,7 +1160,7 @@ extern int drbd_send_protocol(struct drbd_conf *mdev); extern int drbd_send_uuids(struct drbd_conf *mdev); extern int drbd_send_uuids_skip_initial_sync(struct drbd_conf *mdev); extern int drbd_send_sync_uuid(struct drbd_conf *mdev, u64 val); -extern int drbd_send_sizes(struct drbd_conf *mdev, int trigger_reply); +extern int drbd_send_sizes(struct drbd_conf *mdev, int trigger_reply, enum dds_flags flags); extern int _drbd_send_state(struct drbd_conf *mdev); extern int drbd_send_state(struct drbd_conf *mdev); extern int _drbd_send_cmd(struct drbd_conf *mdev, struct socket *sock, @@ -1171,7 +1228,7 @@ extern int drbd_bitmap_io(struct drbd_conf *mdev, int (*io_fn)(struct drbd_conf /* Meta data layout We reserve a 128MB Block (4k aligned) * either at the end of the backing device - * or on a seperate meta data device. */ + * or on a separate meta data device. */ #define MD_RESERVED_SECT (128LU << 11) /* 128 MB, unit sectors */ /* The following numbers are sectors */ @@ -1301,7 +1358,7 @@ struct bm_extent { #define APP_R_HSIZE 15 extern int drbd_bm_init(struct drbd_conf *mdev); -extern int drbd_bm_resize(struct drbd_conf *mdev, sector_t sectors); +extern int drbd_bm_resize(struct drbd_conf *mdev, sector_t sectors, int set_new_bits); extern void drbd_bm_cleanup(struct drbd_conf *mdev); extern void drbd_bm_set_all(struct drbd_conf *mdev); extern void drbd_bm_clear_all(struct drbd_conf *mdev); @@ -1373,7 +1430,7 @@ extern void drbd_resume_io(struct drbd_conf *mdev); extern char *ppsize(char *buf, unsigned long long size); extern sector_t drbd_new_dev_size(struct drbd_conf *, struct drbd_backing_dev *, int); enum determine_dev_size { dev_size_error = -1, unchanged = 0, shrunk = 1, grew = 2 }; -extern enum determine_dev_size drbd_determin_dev_size(struct drbd_conf *, int force) __must_hold(local); +extern enum determine_dev_size drbd_determin_dev_size(struct drbd_conf *, enum dds_flags) __must_hold(local); extern void resync_after_online_grow(struct drbd_conf *); extern void drbd_setup_queue_param(struct drbd_conf *mdev, unsigned int) __must_hold(local); extern int drbd_set_role(struct drbd_conf *mdev, enum drbd_role new_role, @@ -1404,7 +1461,8 @@ static inline void ov_oos_print(struct drbd_conf *mdev) } -extern void drbd_csum(struct drbd_conf *, struct crypto_hash *, struct bio *, void *); +extern void drbd_csum_bio(struct drbd_conf *, struct crypto_hash *, struct bio *, void *); +extern void drbd_csum_ee(struct drbd_conf *, struct crypto_hash *, struct drbd_epoch_entry *, void *); /* worker callbacks */ extern int w_req_cancel_conflict(struct drbd_conf *, struct drbd_work *, int); extern int w_read_retry_remote(struct drbd_conf *, struct drbd_work *, int); @@ -1416,7 +1474,6 @@ extern int w_e_end_ov_req(struct drbd_conf *, struct drbd_work *, int); extern int w_ov_finished(struct drbd_conf *, struct drbd_work *, int); extern int w_resync_inactive(struct drbd_conf *, struct drbd_work *, int); extern int w_resume_next_sg(struct drbd_conf *, struct drbd_work *, int); -extern int w_io_error(struct drbd_conf *, struct drbd_work *, int); extern int w_send_write_hint(struct drbd_conf *, struct drbd_work *, int); extern int w_make_resync_request(struct drbd_conf *, struct drbd_work *, int); extern int w_send_dblock(struct drbd_conf *, struct drbd_work *, int); @@ -1428,6 +1485,8 @@ extern int w_e_reissue(struct drbd_conf *, struct drbd_work *, int); extern void resync_timer_fn(unsigned long data); /* drbd_receiver.c */ +extern int drbd_submit_ee(struct drbd_conf *mdev, struct drbd_epoch_entry *e, + const unsigned rw, const int fault_type); extern int drbd_release_ee(struct drbd_conf *mdev, struct list_head *list); extern struct drbd_epoch_entry *drbd_alloc_ee(struct drbd_conf *mdev, u64 id, @@ -1480,7 +1539,7 @@ static inline void drbd_tcp_nodelay(struct socket *sock) static inline void drbd_tcp_quickack(struct socket *sock) { - int __user val = 1; + int __user val = 2; (void) drbd_setsockopt(sock, SOL_TCP, TCP_QUICKACK, (char __user *)&val, sizeof(val)); } @@ -1583,6 +1642,41 @@ void drbd_bcast_ee(struct drbd_conf *mdev, * inline helper functions *************************/ +/* see also page_chain_add and friends in drbd_receiver.c */ +static inline struct page *page_chain_next(struct page *page) +{ + return (struct page *)page_private(page); +} +#define page_chain_for_each(page) \ + for (; page && ({ prefetch(page_chain_next(page)); 1; }); \ + page = page_chain_next(page)) +#define page_chain_for_each_safe(page, n) \ + for (; page && ({ n = page_chain_next(page); 1; }); page = n) + +static inline int drbd_bio_has_active_page(struct bio *bio) +{ + struct bio_vec *bvec; + int i; + + __bio_for_each_segment(bvec, bio, i, 0) { + if (page_count(bvec->bv_page) > 1) + return 1; + } + + return 0; +} + +static inline int drbd_ee_has_active_page(struct drbd_epoch_entry *e) +{ + struct page *page = e->pages; + page_chain_for_each(page) { + if (page_count(page) > 1) + return 1; + } + return 0; +} + + static inline void drbd_state_lock(struct drbd_conf *mdev) { wait_event(mdev->misc_wait, @@ -1631,7 +1725,7 @@ static inline void __drbd_chk_io_error_(struct drbd_conf *mdev, int forcedetach, switch (mdev->ldev->dc.on_io_error) { case EP_PASS_ON: if (!forcedetach) { - if (printk_ratelimit()) + if (__ratelimit(&drbd_ratelimit_state)) dev_err(DEV, "Local IO failed in %s." "Passing error on...\n", where); break; @@ -2128,7 +2222,7 @@ static inline int __inc_ap_bio_cond(struct drbd_conf *mdev) /* I'd like to use wait_event_lock_irq, * but I'm not sure when it got introduced, * and not sure when it has 3 or 4 arguments */ -static inline void inc_ap_bio(struct drbd_conf *mdev, int one_or_two) +static inline void inc_ap_bio(struct drbd_conf *mdev, int count) { /* compare with after_state_ch, * os.conn != C_WF_BITMAP_S && ns.conn == C_WF_BITMAP_S */ @@ -2150,7 +2244,7 @@ static inline void inc_ap_bio(struct drbd_conf *mdev, int one_or_two) finish_wait(&mdev->misc_wait, &wait); spin_lock_irq(&mdev->req_lock); } - atomic_add(one_or_two, &mdev->ap_bio_cnt); + atomic_add(count, &mdev->ap_bio_cnt); spin_unlock_irq(&mdev->req_lock); } @@ -2241,7 +2335,8 @@ static inline void drbd_md_flush(struct drbd_conf *mdev) if (test_bit(MD_NO_BARRIER, &mdev->flags)) return; - r = blkdev_issue_flush(mdev->ldev->md_bdev, NULL); + r = blkdev_issue_flush(mdev->ldev->md_bdev, GFP_KERNEL, NULL, + BLKDEV_IFL_WAIT); if (r) { set_bit(MD_NO_BARRIER, &mdev->flags); dev_err(DEV, "meta data flush failed with status %d, disabling md-flushes\n", r); diff --git a/drivers/block/drbd/drbd_main.c b/drivers/block/drbd/drbd_main.c index ab871e00ffc5..7258c95e895e 100644 --- a/drivers/block/drbd/drbd_main.c +++ b/drivers/block/drbd/drbd_main.c @@ -684,6 +684,9 @@ static int is_valid_state(struct drbd_conf *mdev, union drbd_state ns) else if (ns.conn > C_CONNECTED && ns.pdsk < D_INCONSISTENT) rv = SS_NO_REMOTE_DISK; + else if (ns.conn > C_CONNECTED && ns.disk < D_UP_TO_DATE && ns.pdsk < D_UP_TO_DATE) + rv = SS_NO_UP_TO_DATE_DISK; + else if ((ns.conn == C_CONNECTED || ns.conn == C_WF_BITMAP_S || ns.conn == C_SYNC_SOURCE || @@ -840,7 +843,12 @@ static union drbd_state sanitize_state(struct drbd_conf *mdev, union drbd_state break; case C_WF_BITMAP_S: case C_PAUSED_SYNC_S: - ns.pdsk = D_OUTDATED; + /* remap any consistent state to D_OUTDATED, + * but disallow "upgrade" of not even consistent states. + */ + ns.pdsk = + (D_DISKLESS < os.pdsk && os.pdsk < D_OUTDATED) + ? os.pdsk : D_OUTDATED; break; case C_SYNC_SOURCE: ns.pdsk = D_INCONSISTENT; @@ -1205,8 +1213,6 @@ static void after_state_ch(struct drbd_conf *mdev, union drbd_state os, && (ns.pdsk < D_INCONSISTENT || ns.pdsk == D_UNKNOWN || ns.pdsk == D_OUTDATED)) { - kfree(mdev->p_uuid); - mdev->p_uuid = NULL; if (get_ldev(mdev)) { if ((ns.role == R_PRIMARY || ns.peer == R_PRIMARY) && mdev->ldev->md.uuid[UI_BITMAP] == 0 && ns.disk >= D_UP_TO_DATE) { @@ -1230,9 +1236,7 @@ static void after_state_ch(struct drbd_conf *mdev, union drbd_state os, /* Last part of the attaching process ... */ if (ns.conn >= C_CONNECTED && os.disk == D_ATTACHING && ns.disk == D_NEGOTIATING) { - kfree(mdev->p_uuid); /* We expect to receive up-to-date UUIDs soon. */ - mdev->p_uuid = NULL; /* ...to not use the old ones in the mean time */ - drbd_send_sizes(mdev, 0); /* to start sync... */ + drbd_send_sizes(mdev, 0, 0); /* to start sync... */ drbd_send_uuids(mdev); drbd_send_state(mdev); } @@ -1668,7 +1672,7 @@ int drbd_send_sync_param(struct drbd_conf *mdev, struct syncer_conf *sc) int drbd_send_protocol(struct drbd_conf *mdev) { struct p_protocol *p; - int size, rv; + int size, cf, rv; size = sizeof(struct p_protocol); @@ -1685,9 +1689,22 @@ int drbd_send_protocol(struct drbd_conf *mdev) p->after_sb_0p = cpu_to_be32(mdev->net_conf->after_sb_0p); p->after_sb_1p = cpu_to_be32(mdev->net_conf->after_sb_1p); p->after_sb_2p = cpu_to_be32(mdev->net_conf->after_sb_2p); - p->want_lose = cpu_to_be32(mdev->net_conf->want_lose); p->two_primaries = cpu_to_be32(mdev->net_conf->two_primaries); + cf = 0; + if (mdev->net_conf->want_lose) + cf |= CF_WANT_LOSE; + if (mdev->net_conf->dry_run) { + if (mdev->agreed_pro_version >= 92) + cf |= CF_DRY_RUN; + else { + dev_err(DEV, "--dry-run is not supported by peer"); + kfree(p); + return 0; + } + } + p->conn_flags = cpu_to_be32(cf); + if (mdev->agreed_pro_version >= 87) strcpy(p->integrity_alg, mdev->net_conf->integrity_alg); @@ -1742,7 +1759,7 @@ int drbd_send_sync_uuid(struct drbd_conf *mdev, u64 val) (struct p_header *)&p, sizeof(p)); } -int drbd_send_sizes(struct drbd_conf *mdev, int trigger_reply) +int drbd_send_sizes(struct drbd_conf *mdev, int trigger_reply, enum dds_flags flags) { struct p_sizes p; sector_t d_size, u_size; @@ -1754,7 +1771,6 @@ int drbd_send_sizes(struct drbd_conf *mdev, int trigger_reply) d_size = drbd_get_max_capacity(mdev->ldev); u_size = mdev->ldev->dc.disk_size; q_order_type = drbd_queue_order_type(mdev); - p.queue_order_type = cpu_to_be32(drbd_queue_order_type(mdev)); put_ldev(mdev); } else { d_size = 0; @@ -1766,7 +1782,8 @@ int drbd_send_sizes(struct drbd_conf *mdev, int trigger_reply) p.u_size = cpu_to_be64(u_size); p.c_size = cpu_to_be64(trigger_reply ? 0 : drbd_get_capacity(mdev->this_bdev)); p.max_segment_size = cpu_to_be32(queue_max_segment_size(mdev->rq_queue)); - p.queue_order_type = cpu_to_be32(q_order_type); + p.queue_order_type = cpu_to_be16(q_order_type); + p.dds_flags = cpu_to_be16(flags); ok = drbd_send_cmd(mdev, USE_DATA_SOCKET, P_SIZES, (struct p_header *)&p, sizeof(p)); @@ -2167,6 +2184,43 @@ int drbd_send_ov_request(struct drbd_conf *mdev, sector_t sector, int size) return ok; } +static int drbd_send_delay_probe(struct drbd_conf *mdev, struct drbd_socket *ds) +{ + struct p_delay_probe dp; + int offset, ok = 0; + struct timeval now; + + mutex_lock(&ds->mutex); + if (likely(ds->socket)) { + do_gettimeofday(&now); + offset = now.tv_usec - mdev->dps_time.tv_usec + + (now.tv_sec - mdev->dps_time.tv_sec) * 1000000; + dp.seq_num = cpu_to_be32(mdev->delay_seq); + dp.offset = cpu_to_be32(offset); + + ok = _drbd_send_cmd(mdev, ds->socket, P_DELAY_PROBE, + (struct p_header *)&dp, sizeof(dp), 0); + } + mutex_unlock(&ds->mutex); + + return ok; +} + +static int drbd_send_delay_probes(struct drbd_conf *mdev) +{ + int ok; + + mdev->delay_seq++; + do_gettimeofday(&mdev->dps_time); + ok = drbd_send_delay_probe(mdev, &mdev->meta); + ok = ok && drbd_send_delay_probe(mdev, &mdev->data); + + mdev->dp_volume_last = mdev->send_cnt; + mod_timer(&mdev->delay_probe_timer, jiffies + mdev->sync_conf.dp_interval * HZ / 10); + + return ok; +} + /* called on sndtimeo * returns FALSE if we should retry, * TRUE if we think connection is dead @@ -2216,9 +2270,9 @@ static int we_should_drop_the_connection(struct drbd_conf *mdev, struct socket * * with page_count == 0 or PageSlab. */ static int _drbd_no_send_page(struct drbd_conf *mdev, struct page *page, - int offset, size_t size) + int offset, size_t size, unsigned msg_flags) { - int sent = drbd_send(mdev, mdev->data.socket, kmap(page) + offset, size, 0); + int sent = drbd_send(mdev, mdev->data.socket, kmap(page) + offset, size, msg_flags); kunmap(page); if (sent == size) mdev->send_cnt += size>>9; @@ -2226,7 +2280,7 @@ static int _drbd_no_send_page(struct drbd_conf *mdev, struct page *page, } static int _drbd_send_page(struct drbd_conf *mdev, struct page *page, - int offset, size_t size) + int offset, size_t size, unsigned msg_flags) { mm_segment_t oldfs = get_fs(); int sent, ok; @@ -2239,14 +2293,15 @@ static int _drbd_send_page(struct drbd_conf *mdev, struct page *page, * __page_cache_release a page that would actually still be referenced * by someone, leading to some obscure delayed Oops somewhere else. */ if (disable_sendpage || (page_count(page) < 1) || PageSlab(page)) - return _drbd_no_send_page(mdev, page, offset, size); + return _drbd_no_send_page(mdev, page, offset, size, msg_flags); + msg_flags |= MSG_NOSIGNAL; drbd_update_congested(mdev); set_fs(KERNEL_DS); do { sent = mdev->data.socket->ops->sendpage(mdev->data.socket, page, offset, len, - MSG_NOSIGNAL); + msg_flags); if (sent == -EAGAIN) { if (we_should_drop_the_connection(mdev, mdev->data.socket)) @@ -2275,9 +2330,11 @@ static int _drbd_send_bio(struct drbd_conf *mdev, struct bio *bio) { struct bio_vec *bvec; int i; + /* hint all but last page with MSG_MORE */ __bio_for_each_segment(bvec, bio, i, 0) { if (!_drbd_no_send_page(mdev, bvec->bv_page, - bvec->bv_offset, bvec->bv_len)) + bvec->bv_offset, bvec->bv_len, + i == bio->bi_vcnt -1 ? 0 : MSG_MORE)) return 0; } return 1; @@ -2287,15 +2344,56 @@ static int _drbd_send_zc_bio(struct drbd_conf *mdev, struct bio *bio) { struct bio_vec *bvec; int i; + /* hint all but last page with MSG_MORE */ __bio_for_each_segment(bvec, bio, i, 0) { if (!_drbd_send_page(mdev, bvec->bv_page, - bvec->bv_offset, bvec->bv_len)) + bvec->bv_offset, bvec->bv_len, + i == bio->bi_vcnt -1 ? 0 : MSG_MORE)) return 0; } + return 1; +} +static int _drbd_send_zc_ee(struct drbd_conf *mdev, struct drbd_epoch_entry *e) +{ + struct page *page = e->pages; + unsigned len = e->size; + /* hint all but last page with MSG_MORE */ + page_chain_for_each(page) { + unsigned l = min_t(unsigned, len, PAGE_SIZE); + if (!_drbd_send_page(mdev, page, 0, l, + page_chain_next(page) ? MSG_MORE : 0)) + return 0; + len -= l; + } return 1; } +static void consider_delay_probes(struct drbd_conf *mdev) +{ + if (mdev->state.conn != C_SYNC_SOURCE || mdev->agreed_pro_version < 93) + return; + + if (mdev->dp_volume_last + mdev->sync_conf.dp_volume * 2 < mdev->send_cnt) + drbd_send_delay_probes(mdev); +} + +static int w_delay_probes(struct drbd_conf *mdev, struct drbd_work *w, int cancel) +{ + if (!cancel && mdev->state.conn == C_SYNC_SOURCE) + drbd_send_delay_probes(mdev); + + return 1; +} + +static void delay_probe_timer_fn(unsigned long data) +{ + struct drbd_conf *mdev = (struct drbd_conf *) data; + + if (list_empty(&mdev->delay_probe_work.list)) + drbd_queue_work(&mdev->data.work, &mdev->delay_probe_work); +} + /* Used to send write requests * R_PRIMARY -> Peer (P_DATA) */ @@ -2344,11 +2442,11 @@ int drbd_send_dblock(struct drbd_conf *mdev, struct drbd_request *req) p.dp_flags = cpu_to_be32(dp_flags); set_bit(UNPLUG_REMOTE, &mdev->flags); ok = (sizeof(p) == - drbd_send(mdev, mdev->data.socket, &p, sizeof(p), MSG_MORE)); + drbd_send(mdev, mdev->data.socket, &p, sizeof(p), dgs ? MSG_MORE : 0)); if (ok && dgs) { dgb = mdev->int_dig_out; - drbd_csum(mdev, mdev->integrity_w_tfm, req->master_bio, dgb); - ok = drbd_send(mdev, mdev->data.socket, dgb, dgs, MSG_MORE); + drbd_csum_bio(mdev, mdev->integrity_w_tfm, req->master_bio, dgb); + ok = drbd_send(mdev, mdev->data.socket, dgb, dgs, 0); } if (ok) { if (mdev->net_conf->wire_protocol == DRBD_PROT_A) @@ -2358,6 +2456,10 @@ int drbd_send_dblock(struct drbd_conf *mdev, struct drbd_request *req) } drbd_put_data_sock(mdev); + + if (ok) + consider_delay_probes(mdev); + return ok; } @@ -2393,16 +2495,20 @@ int drbd_send_block(struct drbd_conf *mdev, enum drbd_packets cmd, return 0; ok = sizeof(p) == drbd_send(mdev, mdev->data.socket, &p, - sizeof(p), MSG_MORE); + sizeof(p), dgs ? MSG_MORE : 0); if (ok && dgs) { dgb = mdev->int_dig_out; - drbd_csum(mdev, mdev->integrity_w_tfm, e->private_bio, dgb); - ok = drbd_send(mdev, mdev->data.socket, dgb, dgs, MSG_MORE); + drbd_csum_ee(mdev, mdev->integrity_w_tfm, e, dgb); + ok = drbd_send(mdev, mdev->data.socket, dgb, dgs, 0); } if (ok) - ok = _drbd_send_zc_bio(mdev, e->private_bio); + ok = _drbd_send_zc_ee(mdev, e); drbd_put_data_sock(mdev); + + if (ok) + consider_delay_probes(mdev); + return ok; } @@ -2615,16 +2721,24 @@ void drbd_init_set_defaults(struct drbd_conf *mdev) INIT_LIST_HEAD(&mdev->unplug_work.list); INIT_LIST_HEAD(&mdev->md_sync_work.list); INIT_LIST_HEAD(&mdev->bm_io_work.w.list); + INIT_LIST_HEAD(&mdev->delay_probes); + INIT_LIST_HEAD(&mdev->delay_probe_work.list); + mdev->resync_work.cb = w_resync_inactive; mdev->unplug_work.cb = w_send_write_hint; mdev->md_sync_work.cb = w_md_sync; mdev->bm_io_work.w.cb = w_bitmap_io; + mdev->delay_probe_work.cb = w_delay_probes; init_timer(&mdev->resync_timer); init_timer(&mdev->md_sync_timer); + init_timer(&mdev->delay_probe_timer); mdev->resync_timer.function = resync_timer_fn; mdev->resync_timer.data = (unsigned long) mdev; mdev->md_sync_timer.function = md_sync_timer_fn; mdev->md_sync_timer.data = (unsigned long) mdev; + mdev->delay_probe_timer.function = delay_probe_timer_fn; + mdev->delay_probe_timer.data = (unsigned long) mdev; + init_waitqueue_head(&mdev->misc_wait); init_waitqueue_head(&mdev->state_wait); @@ -2667,7 +2781,7 @@ void drbd_mdev_cleanup(struct drbd_conf *mdev) drbd_set_my_capacity(mdev, 0); if (mdev->bitmap) { /* maybe never allocated. */ - drbd_bm_resize(mdev, 0); + drbd_bm_resize(mdev, 0, 1); drbd_bm_cleanup(mdev); } @@ -3116,7 +3230,7 @@ int __init drbd_init(void) if (err) goto Enomem; - drbd_proc = proc_create("drbd", S_IFREG | S_IRUGO , NULL, &drbd_proc_fops); + drbd_proc = proc_create_data("drbd", S_IFREG | S_IRUGO , NULL, &drbd_proc_fops, NULL); if (!drbd_proc) { printk(KERN_ERR "drbd: unable to register proc file\n"); goto Enomem; @@ -3161,14 +3275,18 @@ void drbd_free_bc(struct drbd_backing_dev *ldev) void drbd_free_sock(struct drbd_conf *mdev) { if (mdev->data.socket) { + mutex_lock(&mdev->data.mutex); kernel_sock_shutdown(mdev->data.socket, SHUT_RDWR); sock_release(mdev->data.socket); mdev->data.socket = NULL; + mutex_unlock(&mdev->data.mutex); } if (mdev->meta.socket) { + mutex_lock(&mdev->meta.mutex); kernel_sock_shutdown(mdev->meta.socket, SHUT_RDWR); sock_release(mdev->meta.socket); mdev->meta.socket = NULL; + mutex_unlock(&mdev->meta.mutex); } } @@ -3643,7 +3761,8 @@ _drbd_fault_str(unsigned int type) { [DRBD_FAULT_DT_RD] = "Data read", [DRBD_FAULT_DT_RA] = "Data read ahead", [DRBD_FAULT_BM_ALLOC] = "BM allocation", - [DRBD_FAULT_AL_EE] = "EE allocation" + [DRBD_FAULT_AL_EE] = "EE allocation", + [DRBD_FAULT_RECEIVE] = "receive data corruption", }; return (type < DRBD_FAULT_MAX) ? _faults[type] : "**Unknown**"; @@ -3662,7 +3781,7 @@ _drbd_insert_fault(struct drbd_conf *mdev, unsigned int type) if (ret) { fault_count++; - if (printk_ratelimit()) + if (__ratelimit(&drbd_ratelimit_state)) dev_warn(DEV, "***Simulating %s failure\n", _drbd_fault_str(type)); } diff --git a/drivers/block/drbd/drbd_nl.c b/drivers/block/drbd/drbd_nl.c index 1292e0620663..2151f18b21de 100644 --- a/drivers/block/drbd/drbd_nl.c +++ b/drivers/block/drbd/drbd_nl.c @@ -285,8 +285,8 @@ int drbd_set_role(struct drbd_conf *mdev, enum drbd_role new_role, int force) } if (r == SS_NO_UP_TO_DATE_DISK && force && - (mdev->state.disk == D_INCONSISTENT || - mdev->state.disk == D_OUTDATED)) { + (mdev->state.disk < D_UP_TO_DATE && + mdev->state.disk >= D_INCONSISTENT)) { mask.disk = D_MASK; val.disk = D_UP_TO_DATE; forced = 1; @@ -407,7 +407,7 @@ static int drbd_nl_primary(struct drbd_conf *mdev, struct drbd_nl_cfg_req *nlp, } reply->ret_code = - drbd_set_role(mdev, R_PRIMARY, primary_args.overwrite_peer); + drbd_set_role(mdev, R_PRIMARY, primary_args.primary_force); return 0; } @@ -510,7 +510,7 @@ void drbd_resume_io(struct drbd_conf *mdev) * Returns 0 on success, negative return values indicate errors. * You should call drbd_md_sync() after calling this function. */ -enum determine_dev_size drbd_determin_dev_size(struct drbd_conf *mdev, int force) __must_hold(local) +enum determine_dev_size drbd_determin_dev_size(struct drbd_conf *mdev, enum dds_flags flags) __must_hold(local) { sector_t prev_first_sect, prev_size; /* previous meta location */ sector_t la_size; @@ -541,12 +541,12 @@ enum determine_dev_size drbd_determin_dev_size(struct drbd_conf *mdev, int force /* TODO: should only be some assert here, not (re)init... */ drbd_md_set_sector_offsets(mdev, mdev->ldev); - size = drbd_new_dev_size(mdev, mdev->ldev, force); + size = drbd_new_dev_size(mdev, mdev->ldev, flags & DDSF_FORCED); if (drbd_get_capacity(mdev->this_bdev) != size || drbd_bm_capacity(mdev) != size) { int err; - err = drbd_bm_resize(mdev, size); + err = drbd_bm_resize(mdev, size, !(flags & DDSF_NO_RESYNC)); if (unlikely(err)) { /* currently there is only one error: ENOMEM! */ size = drbd_bm_capacity(mdev)>>1; @@ -704,14 +704,10 @@ void drbd_setup_queue_param(struct drbd_conf *mdev, unsigned int max_seg_s) __mu struct request_queue * const b = mdev->ldev->backing_bdev->bd_disk->queue; int max_segments = mdev->ldev->dc.max_bio_bvecs; - if (b->merge_bvec_fn && !mdev->ldev->dc.use_bmbv) - max_seg_s = PAGE_SIZE; - max_seg_s = min(queue_max_sectors(b) * queue_logical_block_size(b), max_seg_s); - blk_queue_max_sectors(q, max_seg_s >> 9); - blk_queue_max_phys_segments(q, max_segments ? max_segments : MAX_PHYS_SEGMENTS); - blk_queue_max_hw_segments(q, max_segments ? max_segments : MAX_HW_SEGMENTS); + blk_queue_max_hw_sectors(q, max_seg_s >> 9); + blk_queue_max_segments(q, max_segments ? max_segments : BLK_MAX_SEGMENTS); blk_queue_max_segment_size(q, max_seg_s); blk_queue_logical_block_size(q, 512); blk_queue_segment_boundary(q, PAGE_SIZE-1); @@ -942,6 +938,25 @@ static int drbd_nl_disk_conf(struct drbd_conf *mdev, struct drbd_nl_cfg_req *nlp drbd_md_set_sector_offsets(mdev, nbc); + /* allocate a second IO page if logical_block_size != 512 */ + logical_block_size = bdev_logical_block_size(nbc->md_bdev); + if (logical_block_size == 0) + logical_block_size = MD_SECTOR_SIZE; + + if (logical_block_size != MD_SECTOR_SIZE) { + if (!mdev->md_io_tmpp) { + struct page *page = alloc_page(GFP_NOIO); + if (!page) + goto force_diskless_dec; + + dev_warn(DEV, "Meta data's bdev logical_block_size = %d != %d\n", + logical_block_size, MD_SECTOR_SIZE); + dev_warn(DEV, "Workaround engaged (has performance impact).\n"); + + mdev->md_io_tmpp = page; + } + } + if (!mdev->bitmap) { if (drbd_bm_init(mdev)) { retcode = ERR_NOMEM; @@ -981,25 +996,6 @@ static int drbd_nl_disk_conf(struct drbd_conf *mdev, struct drbd_nl_cfg_req *nlp goto force_diskless_dec; } - /* allocate a second IO page if logical_block_size != 512 */ - logical_block_size = bdev_logical_block_size(nbc->md_bdev); - if (logical_block_size == 0) - logical_block_size = MD_SECTOR_SIZE; - - if (logical_block_size != MD_SECTOR_SIZE) { - if (!mdev->md_io_tmpp) { - struct page *page = alloc_page(GFP_NOIO); - if (!page) - goto force_diskless_dec; - - dev_warn(DEV, "Meta data's bdev logical_block_size = %d != %d\n", - logical_block_size, MD_SECTOR_SIZE); - dev_warn(DEV, "Workaround engaged (has performance impact).\n"); - - mdev->md_io_tmpp = page; - } - } - /* Reset the "barriers don't work" bits here, then force meta data to * be written, to ensure we determine if barriers are supported. */ if (nbc->dc.no_md_flush) @@ -1118,6 +1114,12 @@ static int drbd_nl_disk_conf(struct drbd_conf *mdev, struct drbd_nl_cfg_req *nlp mdev->new_state_tmp.i = ns.i; ns.i = os.i; ns.disk = D_NEGOTIATING; + + /* We expect to receive up-to-date UUIDs soon. + To avoid a race in receive_state, free p_uuid while + holding req_lock. I.e. atomic with the state change */ + kfree(mdev->p_uuid); + mdev->p_uuid = NULL; } rv = _drbd_set_state(mdev, ns, CS_VERBOSE, NULL); @@ -1200,13 +1202,12 @@ static int drbd_nl_net_conf(struct drbd_conf *mdev, struct drbd_nl_cfg_req *nlp, } /* allocation not in the IO path, cqueue thread context */ - new_conf = kmalloc(sizeof(struct net_conf), GFP_KERNEL); + new_conf = kzalloc(sizeof(struct net_conf), GFP_KERNEL); if (!new_conf) { retcode = ERR_NOMEM; goto fail; } - memset(new_conf, 0, sizeof(struct net_conf)); new_conf->timeout = DRBD_TIMEOUT_DEF; new_conf->try_connect_int = DRBD_CONNECT_INT_DEF; new_conf->ping_int = DRBD_PING_INT_DEF; @@ -1478,8 +1479,8 @@ static int drbd_nl_resize(struct drbd_conf *mdev, struct drbd_nl_cfg_req *nlp, { struct resize rs; int retcode = NO_ERROR; - int ldsc = 0; /* local disk size changed */ enum determine_dev_size dd; + enum dds_flags ddsf; memset(&rs, 0, sizeof(struct resize)); if (!resize_from_tags(mdev, nlp->tag_list, &rs)) { @@ -1503,13 +1504,17 @@ static int drbd_nl_resize(struct drbd_conf *mdev, struct drbd_nl_cfg_req *nlp, goto fail; } - if (mdev->ldev->known_size != drbd_get_capacity(mdev->ldev->backing_bdev)) { - mdev->ldev->known_size = drbd_get_capacity(mdev->ldev->backing_bdev); - ldsc = 1; + if (rs.no_resync && mdev->agreed_pro_version < 93) { + retcode = ERR_NEED_APV_93; + goto fail; } + if (mdev->ldev->known_size != drbd_get_capacity(mdev->ldev->backing_bdev)) + mdev->ldev->known_size = drbd_get_capacity(mdev->ldev->backing_bdev); + mdev->ldev->dc.disk_size = (sector_t)rs.resize_size; - dd = drbd_determin_dev_size(mdev, rs.resize_force); + ddsf = (rs.resize_force ? DDSF_FORCED : 0) | (rs.no_resync ? DDSF_NO_RESYNC : 0); + dd = drbd_determin_dev_size(mdev, ddsf); drbd_md_sync(mdev); put_ldev(mdev); if (dd == dev_size_error) { @@ -1517,12 +1522,12 @@ static int drbd_nl_resize(struct drbd_conf *mdev, struct drbd_nl_cfg_req *nlp, goto fail; } - if (mdev->state.conn == C_CONNECTED && (dd != unchanged || ldsc)) { + if (mdev->state.conn == C_CONNECTED) { if (dd == grew) set_bit(RESIZE_PENDING, &mdev->flags); drbd_send_uuids(mdev); - drbd_send_sizes(mdev, 1); + drbd_send_sizes(mdev, 1, ddsf); } fail: @@ -1552,6 +1557,10 @@ static int drbd_nl_syncer_conf(struct drbd_conf *mdev, struct drbd_nl_cfg_req *n sc.rate = DRBD_RATE_DEF; sc.after = DRBD_AFTER_DEF; sc.al_extents = DRBD_AL_EXTENTS_DEF; + sc.dp_volume = DRBD_DP_VOLUME_DEF; + sc.dp_interval = DRBD_DP_INTERVAL_DEF; + sc.throttle_th = DRBD_RS_THROTTLE_TH_DEF; + sc.hold_off_th = DRBD_RS_HOLD_OFF_TH_DEF; } else memcpy(&sc, &mdev->sync_conf, sizeof(struct syncer_conf)); @@ -2208,9 +2217,9 @@ void drbd_bcast_ee(struct drbd_conf *mdev, { struct cn_msg *cn_reply; struct drbd_nl_cfg_reply *reply; - struct bio_vec *bvec; unsigned short *tl; - int i; + struct page *page; + unsigned len; if (!e) return; @@ -2248,11 +2257,15 @@ void drbd_bcast_ee(struct drbd_conf *mdev, put_unaligned(T_ee_data, tl++); put_unaligned(e->size, tl++); - __bio_for_each_segment(bvec, e->private_bio, i, 0) { - void *d = kmap(bvec->bv_page); - memcpy(tl, d + bvec->bv_offset, bvec->bv_len); - kunmap(bvec->bv_page); - tl=(unsigned short*)((char*)tl + bvec->bv_len); + len = e->size; + page = e->pages; + page_chain_for_each(page) { + void *d = kmap_atomic(page, KM_USER0); + unsigned l = min_t(unsigned, len, PAGE_SIZE); + memcpy(tl, d, l); + kunmap_atomic(d, KM_USER0); + tl = (unsigned short*)((char*)tl + l); + len -= l; } put_unaligned(TT_END, tl++); /* Close the tag list */ diff --git a/drivers/block/drbd/drbd_proc.c b/drivers/block/drbd/drbd_proc.c index df8ad9660d8f..d0f1767ea4c3 100644 --- a/drivers/block/drbd/drbd_proc.c +++ b/drivers/block/drbd/drbd_proc.c @@ -28,7 +28,6 @@ #include <asm/uaccess.h> #include <linux/fs.h> #include <linux/file.h> -#include <linux/slab.h> #include <linux/proc_fs.h> #include <linux/seq_file.h> #include <linux/drbd.h> @@ -74,14 +73,21 @@ static void drbd_syncer_progress(struct drbd_conf *mdev, struct seq_file *seq) seq_printf(seq, "sync'ed:%3u.%u%% ", res / 10, res % 10); /* if more than 1 GB display in MB */ if (mdev->rs_total > 0x100000L) - seq_printf(seq, "(%lu/%lu)M\n\t", + seq_printf(seq, "(%lu/%lu)M", (unsigned long) Bit2KB(rs_left >> 10), (unsigned long) Bit2KB(mdev->rs_total >> 10)); else - seq_printf(seq, "(%lu/%lu)K\n\t", + seq_printf(seq, "(%lu/%lu)K", (unsigned long) Bit2KB(rs_left), (unsigned long) Bit2KB(mdev->rs_total)); + if (mdev->state.conn == C_SYNC_TARGET) + seq_printf(seq, " queue_delay: %d.%d ms\n\t", + mdev->data_delay / 1000, + (mdev->data_delay % 1000) / 100); + else if (mdev->state.conn == C_SYNC_SOURCE) + seq_printf(seq, " delay_probe: %u\n\t", mdev->delay_seq); + /* see drivers/md/md.c * We do not want to overflow, so the order of operands and * the * 100 / 100 trick are important. We do a +1 to be @@ -129,6 +135,14 @@ static void drbd_syncer_progress(struct drbd_conf *mdev, struct seq_file *seq) else seq_printf(seq, " (%ld)", dbdt); + if (mdev->state.conn == C_SYNC_TARGET) { + if (mdev->c_sync_rate > 1000) + seq_printf(seq, " want: %d,%03d", + mdev->c_sync_rate / 1000, mdev->c_sync_rate % 1000); + else + seq_printf(seq, " want: %d", mdev->c_sync_rate); + } + seq_printf(seq, " K/sec\n"); } diff --git a/drivers/block/drbd/drbd_receiver.c b/drivers/block/drbd/drbd_receiver.c index d065c646b35a..dff48701b84d 100644 --- a/drivers/block/drbd/drbd_receiver.c +++ b/drivers/block/drbd/drbd_receiver.c @@ -42,7 +42,6 @@ #include <linux/unistd.h> #include <linux/vmalloc.h> #include <linux/random.h> -#include <linux/mm.h> #include <linux/string.h> #include <linux/scatterlist.h> #include "drbd_int.h" @@ -80,30 +79,128 @@ static struct drbd_epoch *previous_epoch(struct drbd_conf *mdev, struct drbd_epo #define GFP_TRY (__GFP_HIGHMEM | __GFP_NOWARN) -static struct page *drbd_pp_first_page_or_try_alloc(struct drbd_conf *mdev) +/* + * some helper functions to deal with single linked page lists, + * page->private being our "next" pointer. + */ + +/* If at least n pages are linked at head, get n pages off. + * Otherwise, don't modify head, and return NULL. + * Locking is the responsibility of the caller. + */ +static struct page *page_chain_del(struct page **head, int n) +{ + struct page *page; + struct page *tmp; + + BUG_ON(!n); + BUG_ON(!head); + + page = *head; + + if (!page) + return NULL; + + while (page) { + tmp = page_chain_next(page); + if (--n == 0) + break; /* found sufficient pages */ + if (tmp == NULL) + /* insufficient pages, don't use any of them. */ + return NULL; + page = tmp; + } + + /* add end of list marker for the returned list */ + set_page_private(page, 0); + /* actual return value, and adjustment of head */ + page = *head; + *head = tmp; + return page; +} + +/* may be used outside of locks to find the tail of a (usually short) + * "private" page chain, before adding it back to a global chain head + * with page_chain_add() under a spinlock. */ +static struct page *page_chain_tail(struct page *page, int *len) +{ + struct page *tmp; + int i = 1; + while ((tmp = page_chain_next(page))) + ++i, page = tmp; + if (len) + *len = i; + return page; +} + +static int page_chain_free(struct page *page) +{ + struct page *tmp; + int i = 0; + page_chain_for_each_safe(page, tmp) { + put_page(page); + ++i; + } + return i; +} + +static void page_chain_add(struct page **head, + struct page *chain_first, struct page *chain_last) +{ +#if 1 + struct page *tmp; + tmp = page_chain_tail(chain_first, NULL); + BUG_ON(tmp != chain_last); +#endif + + /* add chain to head */ + set_page_private(chain_last, (unsigned long)*head); + *head = chain_first; +} + +static struct page *drbd_pp_first_pages_or_try_alloc(struct drbd_conf *mdev, int number) { struct page *page = NULL; + struct page *tmp = NULL; + int i = 0; /* Yes, testing drbd_pp_vacant outside the lock is racy. * So what. It saves a spin_lock. */ - if (drbd_pp_vacant > 0) { + if (drbd_pp_vacant >= number) { spin_lock(&drbd_pp_lock); - page = drbd_pp_pool; - if (page) { - drbd_pp_pool = (struct page *)page_private(page); - set_page_private(page, 0); /* just to be polite */ - drbd_pp_vacant--; - } + page = page_chain_del(&drbd_pp_pool, number); + if (page) + drbd_pp_vacant -= number; spin_unlock(&drbd_pp_lock); + if (page) + return page; } + /* GFP_TRY, because we must not cause arbitrary write-out: in a DRBD * "criss-cross" setup, that might cause write-out on some other DRBD, * which in turn might block on the other node at this very place. */ - if (!page) - page = alloc_page(GFP_TRY); - if (page) - atomic_inc(&mdev->pp_in_use); - return page; + for (i = 0; i < number; i++) { + tmp = alloc_page(GFP_TRY); + if (!tmp) + break; + set_page_private(tmp, (unsigned long)page); + page = tmp; + } + + if (i == number) + return page; + + /* Not enough pages immediately available this time. + * No need to jump around here, drbd_pp_alloc will retry this + * function "soon". */ + if (page) { + tmp = page_chain_tail(page, NULL); + spin_lock(&drbd_pp_lock); + page_chain_add(&drbd_pp_pool, page, tmp); + drbd_pp_vacant += i; + spin_unlock(&drbd_pp_lock); + } + return NULL; } /* kick lower level device, if we have more than (arbitrary number) @@ -127,7 +224,7 @@ static void reclaim_net_ee(struct drbd_conf *mdev, struct list_head *to_be_freed list_for_each_safe(le, tle, &mdev->net_ee) { e = list_entry(le, struct drbd_epoch_entry, w.list); - if (drbd_bio_has_active_page(e->private_bio)) + if (drbd_ee_has_active_page(e)) break; list_move(le, to_be_freed); } @@ -148,32 +245,34 @@ static void drbd_kick_lo_and_reclaim_net(struct drbd_conf *mdev) } /** - * drbd_pp_alloc() - Returns a page, fails only if a signal comes in + * drbd_pp_alloc() - Returns @number pages, retries forever (or until signalled) * @mdev: DRBD device. - * @retry: whether or not to retry allocation forever (or until signalled) + * @number: number of pages requested + * @retry: whether to retry, if not enough pages are available right now * - * Tries to allocate a page, first from our own page pool, then from the - * kernel, unless this allocation would exceed the max_buffers setting. - * If @retry is non-zero, retry until DRBD frees a page somewhere else. + * Tries to allocate number pages, first from our own page pool, then from + * the kernel, unless this allocation would exceed the max_buffers setting. + * Possibly retry until DRBD frees sufficient pages somewhere else. + * + * Returns a page chain linked via page->private. */ -static struct page *drbd_pp_alloc(struct drbd_conf *mdev, int retry) +static struct page *drbd_pp_alloc(struct drbd_conf *mdev, unsigned number, bool retry) { struct page *page = NULL; DEFINE_WAIT(wait); - if (atomic_read(&mdev->pp_in_use) < mdev->net_conf->max_buffers) { - page = drbd_pp_first_page_or_try_alloc(mdev); - if (page) - return page; - } + /* Yes, we may run up to @number over max_buffers. If we + * follow it strictly, the admin will get it wrong anyways. */ + if (atomic_read(&mdev->pp_in_use) < mdev->net_conf->max_buffers) + page = drbd_pp_first_pages_or_try_alloc(mdev, number); - for (;;) { + while (page == NULL) { prepare_to_wait(&drbd_pp_wait, &wait, TASK_INTERRUPTIBLE); drbd_kick_lo_and_reclaim_net(mdev); if (atomic_read(&mdev->pp_in_use) < mdev->net_conf->max_buffers) { - page = drbd_pp_first_page_or_try_alloc(mdev); + page = drbd_pp_first_pages_or_try_alloc(mdev, number); if (page) break; } @@ -190,62 +289,32 @@ static struct page *drbd_pp_alloc(struct drbd_conf *mdev, int retry) } finish_wait(&drbd_pp_wait, &wait); + if (page) + atomic_add(number, &mdev->pp_in_use); return page; } /* Must not be used from irq, as that may deadlock: see drbd_pp_alloc. - * Is also used from inside an other spin_lock_irq(&mdev->req_lock) */ + * Is also used from inside an other spin_lock_irq(&mdev->req_lock); + * Either links the page chain back to the global pool, + * or returns all pages to the system. */ static void drbd_pp_free(struct drbd_conf *mdev, struct page *page) { - int free_it; - - spin_lock(&drbd_pp_lock); - if (drbd_pp_vacant > (DRBD_MAX_SEGMENT_SIZE/PAGE_SIZE)*minor_count) { - free_it = 1; - } else { - set_page_private(page, (unsigned long)drbd_pp_pool); - drbd_pp_pool = page; - drbd_pp_vacant++; - free_it = 0; - } - spin_unlock(&drbd_pp_lock); - - atomic_dec(&mdev->pp_in_use); - - if (free_it) - __free_page(page); - - wake_up(&drbd_pp_wait); -} - -static void drbd_pp_free_bio_pages(struct drbd_conf *mdev, struct bio *bio) -{ - struct page *p_to_be_freed = NULL; - struct page *page; - struct bio_vec *bvec; int i; - - spin_lock(&drbd_pp_lock); - __bio_for_each_segment(bvec, bio, i, 0) { - if (drbd_pp_vacant > (DRBD_MAX_SEGMENT_SIZE/PAGE_SIZE)*minor_count) { - set_page_private(bvec->bv_page, (unsigned long)p_to_be_freed); - p_to_be_freed = bvec->bv_page; - } else { - set_page_private(bvec->bv_page, (unsigned long)drbd_pp_pool); - drbd_pp_pool = bvec->bv_page; - drbd_pp_vacant++; - } - } - spin_unlock(&drbd_pp_lock); - atomic_sub(bio->bi_vcnt, &mdev->pp_in_use); - - while (p_to_be_freed) { - page = p_to_be_freed; - p_to_be_freed = (struct page *)page_private(page); - set_page_private(page, 0); /* just to be polite */ - put_page(page); + if (drbd_pp_vacant > (DRBD_MAX_SEGMENT_SIZE/PAGE_SIZE)*minor_count) + i = page_chain_free(page); + else { + struct page *tmp; + tmp = page_chain_tail(page, &i); + spin_lock(&drbd_pp_lock); + page_chain_add(&drbd_pp_pool, page, tmp); + drbd_pp_vacant += i; + spin_unlock(&drbd_pp_lock); } - + atomic_sub(i, &mdev->pp_in_use); + i = atomic_read(&mdev->pp_in_use); + if (i < 0) + dev_warn(DEV, "ASSERTION FAILED: pp_in_use: %d < 0\n", i); wake_up(&drbd_pp_wait); } @@ -270,11 +339,9 @@ struct drbd_epoch_entry *drbd_alloc_ee(struct drbd_conf *mdev, unsigned int data_size, gfp_t gfp_mask) __must_hold(local) { - struct request_queue *q; struct drbd_epoch_entry *e; struct page *page; - struct bio *bio; - unsigned int ds; + unsigned nr_pages = (data_size + PAGE_SIZE -1) >> PAGE_SHIFT; if (FAULT_ACTIVE(mdev, DRBD_FAULT_AL_EE)) return NULL; @@ -286,84 +353,32 @@ struct drbd_epoch_entry *drbd_alloc_ee(struct drbd_conf *mdev, return NULL; } - bio = bio_alloc(gfp_mask & ~__GFP_HIGHMEM, div_ceil(data_size, PAGE_SIZE)); - if (!bio) { - if (!(gfp_mask & __GFP_NOWARN)) - dev_err(DEV, "alloc_ee: Allocation of a bio failed\n"); - goto fail1; - } - - bio->bi_bdev = mdev->ldev->backing_bdev; - bio->bi_sector = sector; - - ds = data_size; - while (ds) { - page = drbd_pp_alloc(mdev, (gfp_mask & __GFP_WAIT)); - if (!page) { - if (!(gfp_mask & __GFP_NOWARN)) - dev_err(DEV, "alloc_ee: Allocation of a page failed\n"); - goto fail2; - } - if (!bio_add_page(bio, page, min_t(int, ds, PAGE_SIZE), 0)) { - drbd_pp_free(mdev, page); - dev_err(DEV, "alloc_ee: bio_add_page(s=%llu," - "data_size=%u,ds=%u) failed\n", - (unsigned long long)sector, data_size, ds); - - q = bdev_get_queue(bio->bi_bdev); - if (q->merge_bvec_fn) { - struct bvec_merge_data bvm = { - .bi_bdev = bio->bi_bdev, - .bi_sector = bio->bi_sector, - .bi_size = bio->bi_size, - .bi_rw = bio->bi_rw, - }; - int l = q->merge_bvec_fn(q, &bvm, - &bio->bi_io_vec[bio->bi_vcnt]); - dev_err(DEV, "merge_bvec_fn() = %d\n", l); - } - - /* dump more of the bio. */ - dev_err(DEV, "bio->bi_max_vecs = %d\n", bio->bi_max_vecs); - dev_err(DEV, "bio->bi_vcnt = %d\n", bio->bi_vcnt); - dev_err(DEV, "bio->bi_size = %d\n", bio->bi_size); - dev_err(DEV, "bio->bi_phys_segments = %d\n", bio->bi_phys_segments); - - goto fail2; - break; - } - ds -= min_t(int, ds, PAGE_SIZE); - } - - D_ASSERT(data_size == bio->bi_size); - - bio->bi_private = e; - e->mdev = mdev; - e->sector = sector; - e->size = bio->bi_size; + page = drbd_pp_alloc(mdev, nr_pages, (gfp_mask & __GFP_WAIT)); + if (!page) + goto fail; - e->private_bio = bio; - e->block_id = id; INIT_HLIST_NODE(&e->colision); e->epoch = NULL; + e->mdev = mdev; + e->pages = page; + atomic_set(&e->pending_bios, 0); + e->size = data_size; e->flags = 0; + e->sector = sector; + e->sector = sector; + e->block_id = id; return e; - fail2: - drbd_pp_free_bio_pages(mdev, bio); - bio_put(bio); - fail1: + fail: mempool_free(e, drbd_ee_mempool); - return NULL; } void drbd_free_ee(struct drbd_conf *mdev, struct drbd_epoch_entry *e) { - struct bio *bio = e->private_bio; - drbd_pp_free_bio_pages(mdev, bio); - bio_put(bio); + drbd_pp_free(mdev, e->pages); + D_ASSERT(atomic_read(&e->pending_bios) == 0); D_ASSERT(hlist_unhashed(&e->colision)); mempool_free(e, drbd_ee_mempool); } @@ -555,6 +570,25 @@ static int drbd_recv(struct drbd_conf *mdev, void *buf, size_t size) return rv; } +/* quoting tcp(7): + * On individual connections, the socket buffer size must be set prior to the + * listen(2) or connect(2) calls in order to have it take effect. + * This is our wrapper to do so. + */ +static void drbd_setbufsize(struct socket *sock, unsigned int snd, + unsigned int rcv) +{ + /* open coded SO_SNDBUF, SO_RCVBUF */ + if (snd) { + sock->sk->sk_sndbuf = snd; + sock->sk->sk_userlocks |= SOCK_SNDBUF_LOCK; + } + if (rcv) { + sock->sk->sk_rcvbuf = rcv; + sock->sk->sk_userlocks |= SOCK_RCVBUF_LOCK; + } +} + static struct socket *drbd_try_connect(struct drbd_conf *mdev) { const char *what; @@ -576,6 +610,8 @@ static struct socket *drbd_try_connect(struct drbd_conf *mdev) sock->sk->sk_rcvtimeo = sock->sk->sk_sndtimeo = mdev->net_conf->try_connect_int*HZ; + drbd_setbufsize(sock, mdev->net_conf->sndbuf_size, + mdev->net_conf->rcvbuf_size); /* explicitly bind to the configured IP as source IP * for the outgoing connections. @@ -654,6 +690,8 @@ static struct socket *drbd_wait_for_connect(struct drbd_conf *mdev) s_listen->sk->sk_reuse = 1; /* SO_REUSEADDR */ s_listen->sk->sk_rcvtimeo = timeo; s_listen->sk->sk_sndtimeo = timeo; + drbd_setbufsize(s_listen, mdev->net_conf->sndbuf_size, + mdev->net_conf->rcvbuf_size); what = "bind before listen"; err = s_listen->ops->bind(s_listen, @@ -840,16 +878,6 @@ retry: sock->sk->sk_priority = TC_PRIO_INTERACTIVE_BULK; msock->sk->sk_priority = TC_PRIO_INTERACTIVE; - if (mdev->net_conf->sndbuf_size) { - sock->sk->sk_sndbuf = mdev->net_conf->sndbuf_size; - sock->sk->sk_userlocks |= SOCK_SNDBUF_LOCK; - } - - if (mdev->net_conf->rcvbuf_size) { - sock->sk->sk_rcvbuf = mdev->net_conf->rcvbuf_size; - sock->sk->sk_userlocks |= SOCK_RCVBUF_LOCK; - } - /* NOT YET ... * sock->sk->sk_sndtimeo = mdev->net_conf->timeout*HZ/10; * sock->sk->sk_rcvtimeo = MAX_SCHEDULE_TIMEOUT; @@ -899,9 +927,10 @@ retry: drbd_thread_start(&mdev->asender); - drbd_send_protocol(mdev); + if (!drbd_send_protocol(mdev)) + return -1; drbd_send_sync_param(mdev, &mdev->sync_conf); - drbd_send_sizes(mdev, 0); + drbd_send_sizes(mdev, 0, 0); drbd_send_uuids(mdev); drbd_send_state(mdev); clear_bit(USE_DEGR_WFC_T, &mdev->flags); @@ -945,7 +974,8 @@ static enum finish_epoch drbd_flush_after_epoch(struct drbd_conf *mdev, struct d int rv; if (mdev->write_ordering >= WO_bdev_flush && get_ldev(mdev)) { - rv = blkdev_issue_flush(mdev->ldev->backing_bdev, NULL); + rv = blkdev_issue_flush(mdev->ldev->backing_bdev, GFP_KERNEL, + NULL, BLKDEV_IFL_WAIT); if (rv) { dev_err(DEV, "local disk flush failed with status %d\n", rv); /* would rather check on EOPNOTSUPP, but that is not reliable. @@ -1119,6 +1149,90 @@ void drbd_bump_write_ordering(struct drbd_conf *mdev, enum write_ordering_e wo) } /** + * drbd_submit_ee() + * @mdev: DRBD device. + * @e: epoch entry + * @rw: flag field, see bio->bi_rw + */ +/* TODO allocate from our own bio_set. */ +int drbd_submit_ee(struct drbd_conf *mdev, struct drbd_epoch_entry *e, + const unsigned rw, const int fault_type) +{ + struct bio *bios = NULL; + struct bio *bio; + struct page *page = e->pages; + sector_t sector = e->sector; + unsigned ds = e->size; + unsigned n_bios = 0; + unsigned nr_pages = (ds + PAGE_SIZE -1) >> PAGE_SHIFT; + + /* In most cases, we will only need one bio. But in case the lower + * level restrictions happen to be different at this offset on this + * side than those of the sending peer, we may need to submit the + * request in more than one bio. */ +next_bio: + bio = bio_alloc(GFP_NOIO, nr_pages); + if (!bio) { + dev_err(DEV, "submit_ee: Allocation of a bio failed\n"); + goto fail; + } + /* > e->sector, unless this is the first bio */ + bio->bi_sector = sector; + bio->bi_bdev = mdev->ldev->backing_bdev; + /* we special case some flags in the multi-bio case, see below + * (BIO_RW_UNPLUG, BIO_RW_BARRIER) */ + bio->bi_rw = rw; + bio->bi_private = e; + bio->bi_end_io = drbd_endio_sec; + + bio->bi_next = bios; + bios = bio; + ++n_bios; + + page_chain_for_each(page) { + unsigned len = min_t(unsigned, ds, PAGE_SIZE); + if (!bio_add_page(bio, page, len, 0)) { + /* a single page must always be possible! */ + BUG_ON(bio->bi_vcnt == 0); + goto next_bio; + } + ds -= len; + sector += len >> 9; + --nr_pages; + } + D_ASSERT(page == NULL); + D_ASSERT(ds == 0); + + atomic_set(&e->pending_bios, n_bios); + do { + bio = bios; + bios = bios->bi_next; + bio->bi_next = NULL; + + /* strip off BIO_RW_UNPLUG unless it is the last bio */ + if (bios) + bio->bi_rw &= ~(1<<BIO_RW_UNPLUG); + + drbd_generic_make_request(mdev, fault_type, bio); + + /* strip off BIO_RW_BARRIER, + * unless it is the first or last bio */ + if (bios && bios->bi_next) + bios->bi_rw &= ~(1<<BIO_RW_BARRIER); + } while (bios); + maybe_kick_lo(mdev); + return 0; + +fail: + while (bios) { + bio = bios; + bios = bios->bi_next; + bio_put(bio); + } + return -ENOMEM; +} + +/** * w_e_reissue() - Worker callback; Resubmit a bio, without BIO_RW_BARRIER set * @mdev: DRBD device. * @w: work object. @@ -1127,8 +1241,6 @@ void drbd_bump_write_ordering(struct drbd_conf *mdev, enum write_ordering_e wo) int w_e_reissue(struct drbd_conf *mdev, struct drbd_work *w, int cancel) __releases(local) { struct drbd_epoch_entry *e = (struct drbd_epoch_entry *)w; - struct bio *bio = e->private_bio; - /* We leave DE_CONTAINS_A_BARRIER and EE_IS_BARRIER in place, (and DE_BARRIER_IN_NEXT_EPOCH_ISSUED in the previous Epoch) so that we can finish that epoch in drbd_may_finish_epoch(). @@ -1142,33 +1254,17 @@ int w_e_reissue(struct drbd_conf *mdev, struct drbd_work *w, int cancel) __relea if (previous_epoch(mdev, e->epoch)) dev_warn(DEV, "Write ordering was not enforced (one time event)\n"); - /* prepare bio for re-submit, - * re-init volatile members */ /* we still have a local reference, * get_ldev was done in receive_Data. */ - bio->bi_bdev = mdev->ldev->backing_bdev; - bio->bi_sector = e->sector; - bio->bi_size = e->size; - bio->bi_idx = 0; - - bio->bi_flags &= ~(BIO_POOL_MASK - 1); - bio->bi_flags |= 1 << BIO_UPTODATE; - - /* don't know whether this is necessary: */ - bio->bi_phys_segments = 0; - bio->bi_next = NULL; - - /* these should be unchanged: */ - /* bio->bi_end_io = drbd_endio_write_sec; */ - /* bio->bi_vcnt = whatever; */ e->w.cb = e_end_block; - - /* This is no longer a barrier request. */ - bio->bi_rw &= ~(1UL << BIO_RW_BARRIER); - - drbd_generic_make_request(mdev, DRBD_FAULT_DT_WR, bio); - + if (drbd_submit_ee(mdev, e, WRITE, DRBD_FAULT_DT_WR) != 0) { + /* drbd_submit_ee fails for one reason only: + * if was not able to allocate sufficient bios. + * requeue, try again later. */ + e->w.cb = w_e_reissue; + drbd_queue_work(&mdev->data.work, &e->w); + } return 1; } @@ -1260,13 +1356,13 @@ static int receive_Barrier(struct drbd_conf *mdev, struct p_header *h) static struct drbd_epoch_entry * read_in_block(struct drbd_conf *mdev, u64 id, sector_t sector, int data_size) __must_hold(local) { + const sector_t capacity = drbd_get_capacity(mdev->this_bdev); struct drbd_epoch_entry *e; - struct bio_vec *bvec; struct page *page; - struct bio *bio; - int dgs, ds, i, rr; + int dgs, ds, rr; void *dig_in = mdev->int_dig_in; void *dig_vv = mdev->int_dig_vv; + unsigned long *data; dgs = (mdev->agreed_pro_version >= 87 && mdev->integrity_r_tfm) ? crypto_hash_digestsize(mdev->integrity_r_tfm) : 0; @@ -1285,29 +1381,44 @@ read_in_block(struct drbd_conf *mdev, u64 id, sector_t sector, int data_size) __ ERR_IF(data_size & 0x1ff) return NULL; ERR_IF(data_size > DRBD_MAX_SEGMENT_SIZE) return NULL; + /* even though we trust out peer, + * we sometimes have to double check. */ + if (sector + (data_size>>9) > capacity) { + dev_err(DEV, "capacity: %llus < sector: %llus + size: %u\n", + (unsigned long long)capacity, + (unsigned long long)sector, data_size); + return NULL; + } + /* GFP_NOIO, because we must not cause arbitrary write-out: in a DRBD * "criss-cross" setup, that might cause write-out on some other DRBD, * which in turn might block on the other node at this very place. */ e = drbd_alloc_ee(mdev, id, sector, data_size, GFP_NOIO); if (!e) return NULL; - bio = e->private_bio; + ds = data_size; - bio_for_each_segment(bvec, bio, i) { - page = bvec->bv_page; - rr = drbd_recv(mdev, kmap(page), min_t(int, ds, PAGE_SIZE)); + page = e->pages; + page_chain_for_each(page) { + unsigned len = min_t(int, ds, PAGE_SIZE); + data = kmap(page); + rr = drbd_recv(mdev, data, len); + if (FAULT_ACTIVE(mdev, DRBD_FAULT_RECEIVE)) { + dev_err(DEV, "Fault injection: Corrupting data on receive\n"); + data[0] = data[0] ^ (unsigned long)-1; + } kunmap(page); - if (rr != min_t(int, ds, PAGE_SIZE)) { + if (rr != len) { drbd_free_ee(mdev, e); dev_warn(DEV, "short read receiving data: read %d expected %d\n", - rr, min_t(int, ds, PAGE_SIZE)); + rr, len); return NULL; } ds -= rr; } if (dgs) { - drbd_csum(mdev, mdev->integrity_r_tfm, bio, dig_vv); + drbd_csum_ee(mdev, mdev->integrity_r_tfm, e, dig_vv); if (memcmp(dig_in, dig_vv, dgs)) { dev_err(DEV, "Digest integrity check FAILED.\n"); drbd_bcast_ee(mdev, "digest failed", @@ -1329,7 +1440,10 @@ static int drbd_drain_block(struct drbd_conf *mdev, int data_size) int rr, rv = 1; void *data; - page = drbd_pp_alloc(mdev, 1); + if (!data_size) + return TRUE; + + page = drbd_pp_alloc(mdev, 1, 1); data = kmap(page); while (data_size) { @@ -1393,7 +1507,7 @@ static int recv_dless_read(struct drbd_conf *mdev, struct drbd_request *req, } if (dgs) { - drbd_csum(mdev, mdev->integrity_r_tfm, bio, dig_vv); + drbd_csum_bio(mdev, mdev->integrity_r_tfm, bio, dig_vv); if (memcmp(dig_in, dig_vv, dgs)) { dev_err(DEV, "Digest integrity check FAILED. Broken NICs?\n"); return 0; @@ -1414,7 +1528,7 @@ static int e_end_resync_block(struct drbd_conf *mdev, struct drbd_work *w, int u D_ASSERT(hlist_unhashed(&e->colision)); - if (likely(drbd_bio_uptodate(e->private_bio))) { + if (likely((e->flags & EE_WAS_ERROR) == 0)) { drbd_set_in_sync(mdev, sector, e->size); ok = drbd_send_ack(mdev, P_RS_WRITE_ACK, e); } else { @@ -1433,30 +1547,28 @@ static int recv_resync_read(struct drbd_conf *mdev, sector_t sector, int data_si struct drbd_epoch_entry *e; e = read_in_block(mdev, ID_SYNCER, sector, data_size); - if (!e) { - put_ldev(mdev); - return FALSE; - } + if (!e) + goto fail; dec_rs_pending(mdev); - e->private_bio->bi_end_io = drbd_endio_write_sec; - e->private_bio->bi_rw = WRITE; - e->w.cb = e_end_resync_block; - inc_unacked(mdev); /* corresponding dec_unacked() in e_end_resync_block() * respective _drbd_clear_done_ee */ + e->w.cb = e_end_resync_block; + spin_lock_irq(&mdev->req_lock); list_add(&e->w.list, &mdev->sync_ee); spin_unlock_irq(&mdev->req_lock); - drbd_generic_make_request(mdev, DRBD_FAULT_RS_WR, e->private_bio); - /* accounting done in endio */ + if (drbd_submit_ee(mdev, e, WRITE, DRBD_FAULT_RS_WR) == 0) + return TRUE; - maybe_kick_lo(mdev); - return TRUE; + drbd_free_ee(mdev, e); +fail: + put_ldev(mdev); + return FALSE; } static int receive_DataReply(struct drbd_conf *mdev, struct p_header *h) @@ -1551,7 +1663,7 @@ static int e_end_block(struct drbd_conf *mdev, struct drbd_work *w, int cancel) } if (mdev->net_conf->wire_protocol == DRBD_PROT_C) { - if (likely(drbd_bio_uptodate(e->private_bio))) { + if (likely((e->flags & EE_WAS_ERROR) == 0)) { pcmd = (mdev->state.conn >= C_SYNC_SOURCE && mdev->state.conn <= C_PAUSED_SYNC_T && e->flags & EE_MAY_SET_IN_SYNC) ? @@ -1697,7 +1809,6 @@ static int receive_Data(struct drbd_conf *mdev, struct p_header *h) return FALSE; } - e->private_bio->bi_end_io = drbd_endio_write_sec; e->w.cb = e_end_block; spin_lock(&mdev->epoch_lock); @@ -1893,12 +2004,8 @@ static int receive_Data(struct drbd_conf *mdev, struct p_header *h) drbd_al_begin_io(mdev, e->sector); } - e->private_bio->bi_rw = rw; - drbd_generic_make_request(mdev, DRBD_FAULT_DT_WR, e->private_bio); - /* accounting done in endio */ - - maybe_kick_lo(mdev); - return TRUE; + if (drbd_submit_ee(mdev, e, rw, DRBD_FAULT_DT_WR) == 0) + return TRUE; out_interrupted: /* yes, the epoch_size now is imbalanced. @@ -1944,7 +2051,7 @@ static int receive_DataRequest(struct drbd_conf *mdev, struct p_header *h) "no local data.\n"); drbd_send_ack_rp(mdev, h->command == P_DATA_REQUEST ? P_NEG_DREPLY : P_NEG_RS_DREPLY , p); - return TRUE; + return drbd_drain_block(mdev, h->length - brps); } /* GFP_NOIO, because we must not cause arbitrary write-out: in a DRBD @@ -1956,9 +2063,6 @@ static int receive_DataRequest(struct drbd_conf *mdev, struct p_header *h) return FALSE; } - e->private_bio->bi_rw = READ; - e->private_bio->bi_end_io = drbd_endio_read_sec; - switch (h->command) { case P_DATA_REQUEST: e->w.cb = w_e_end_data_req; @@ -2052,10 +2156,8 @@ static int receive_DataRequest(struct drbd_conf *mdev, struct p_header *h) inc_unacked(mdev); - drbd_generic_make_request(mdev, fault_type, e->private_bio); - maybe_kick_lo(mdev); - - return TRUE; + if (drbd_submit_ee(mdev, e, READ, fault_type) == 0) + return TRUE; out_free_e: kfree(di); @@ -2472,6 +2574,9 @@ static enum drbd_conns drbd_sync_handshake(struct drbd_conf *mdev, enum drbd_rol hg > 0 ? "source" : "target"); } + if (abs(hg) == 100) + drbd_khelper(mdev, "initial-split-brain"); + if (hg == 100 || (hg == -100 && mdev->net_conf->always_asbp)) { int pcount = (mdev->state.role == R_PRIMARY) + (peer_role == R_PRIMARY); @@ -2513,7 +2618,11 @@ static enum drbd_conns drbd_sync_handshake(struct drbd_conf *mdev, enum drbd_rol } if (hg == -100) { - dev_alert(DEV, "Split-Brain detected, dropping connection!\n"); + /* FIXME this log message is not correct if we end up here + * after an attempted attach on a diskless node. + * We just refuse to attach -- well, we drop the "connection" + * to that disk, in a way... */ + dev_alert(DEV, "Split-Brain detected but unresolved, dropping connection!\n"); drbd_khelper(mdev, "split-brain"); return C_MASK; } @@ -2538,6 +2647,16 @@ static enum drbd_conns drbd_sync_handshake(struct drbd_conf *mdev, enum drbd_rol } } + if (mdev->net_conf->dry_run || test_bit(CONN_DRY_RUN, &mdev->flags)) { + if (hg == 0) + dev_info(DEV, "dry-run connect: No resync, would become Connected immediately.\n"); + else + dev_info(DEV, "dry-run connect: Would become %s, doing a %s resync.", + drbd_conn_str(hg > 0 ? C_SYNC_SOURCE : C_SYNC_TARGET), + abs(hg) >= 2 ? "full" : "bit-map based"); + return C_MASK; + } + if (abs(hg) >= 2) { dev_info(DEV, "Writing the whole bitmap, full sync required after drbd_sync_handshake.\n"); if (drbd_bitmap_io(mdev, &drbd_bmio_set_n_write, "set_n_write from sync_handshake")) @@ -2585,7 +2704,7 @@ static int receive_protocol(struct drbd_conf *mdev, struct p_header *h) struct p_protocol *p = (struct p_protocol *)h; int header_size, data_size; int p_proto, p_after_sb_0p, p_after_sb_1p, p_after_sb_2p; - int p_want_lose, p_two_primaries; + int p_want_lose, p_two_primaries, cf; char p_integrity_alg[SHARED_SECRET_MAX] = ""; header_size = sizeof(*p) - sizeof(*h); @@ -2598,8 +2717,14 @@ static int receive_protocol(struct drbd_conf *mdev, struct p_header *h) p_after_sb_0p = be32_to_cpu(p->after_sb_0p); p_after_sb_1p = be32_to_cpu(p->after_sb_1p); p_after_sb_2p = be32_to_cpu(p->after_sb_2p); - p_want_lose = be32_to_cpu(p->want_lose); p_two_primaries = be32_to_cpu(p->two_primaries); + cf = be32_to_cpu(p->conn_flags); + p_want_lose = cf & CF_WANT_LOSE; + + clear_bit(CONN_DRY_RUN, &mdev->flags); + + if (cf & CF_DRY_RUN) + set_bit(CONN_DRY_RUN, &mdev->flags); if (p_proto != mdev->net_conf->wire_protocol) { dev_err(DEV, "incompatible communication protocols\n"); @@ -2828,7 +2953,7 @@ static int receive_sizes(struct drbd_conf *mdev, struct p_header *h) unsigned int max_seg_s; sector_t p_size, p_usize, my_usize; int ldsc = 0; /* local disk size changed */ - enum drbd_conns nconn; + enum dds_flags ddsf; ERR_IF(h->length != (sizeof(*p)-sizeof(*h))) return FALSE; if (drbd_recv(mdev, h->payload, h->length) != h->length) @@ -2884,8 +3009,9 @@ static int receive_sizes(struct drbd_conf *mdev, struct p_header *h) } #undef min_not_zero + ddsf = be16_to_cpu(p->dds_flags); if (get_ldev(mdev)) { - dd = drbd_determin_dev_size(mdev, 0); + dd = drbd_determin_dev_size(mdev, ddsf); put_ldev(mdev); if (dd == dev_size_error) return FALSE; @@ -2895,33 +3021,21 @@ static int receive_sizes(struct drbd_conf *mdev, struct p_header *h) drbd_set_my_capacity(mdev, p_size); } - if (mdev->p_uuid && mdev->state.conn <= C_CONNECTED && get_ldev(mdev)) { - nconn = drbd_sync_handshake(mdev, - mdev->state.peer, mdev->state.pdsk); - put_ldev(mdev); - - if (nconn == C_MASK) { - drbd_force_state(mdev, NS(conn, C_DISCONNECTING)); - return FALSE; - } - - if (drbd_request_state(mdev, NS(conn, nconn)) < SS_SUCCESS) { - drbd_force_state(mdev, NS(conn, C_DISCONNECTING)); - return FALSE; - } - } - if (get_ldev(mdev)) { if (mdev->ldev->known_size != drbd_get_capacity(mdev->ldev->backing_bdev)) { mdev->ldev->known_size = drbd_get_capacity(mdev->ldev->backing_bdev); ldsc = 1; } - max_seg_s = be32_to_cpu(p->max_segment_size); + if (mdev->agreed_pro_version < 94) + max_seg_s = be32_to_cpu(p->max_segment_size); + else /* drbd 8.3.8 onwards */ + max_seg_s = DRBD_MAX_SEGMENT_SIZE; + if (max_seg_s != queue_max_segment_size(mdev->rq_queue)) drbd_setup_queue_param(mdev, max_seg_s); - drbd_setup_order_type(mdev, be32_to_cpu(p->queue_order_type)); + drbd_setup_order_type(mdev, be16_to_cpu(p->queue_order_type)); put_ldev(mdev); } @@ -2930,14 +3044,17 @@ static int receive_sizes(struct drbd_conf *mdev, struct p_header *h) drbd_get_capacity(mdev->this_bdev) || ldsc) { /* we have different sizes, probably peer * needs to know my new size... */ - drbd_send_sizes(mdev, 0); + drbd_send_sizes(mdev, 0, ddsf); } if (test_and_clear_bit(RESIZE_PENDING, &mdev->flags) || (dd == grew && mdev->state.conn == C_CONNECTED)) { if (mdev->state.pdsk >= D_INCONSISTENT && - mdev->state.disk >= D_INCONSISTENT) - resync_after_online_grow(mdev); - else + mdev->state.disk >= D_INCONSISTENT) { + if (ddsf & DDSF_NO_RESYNC) + dev_info(DEV, "Resync of new storage suppressed with --assume-clean\n"); + else + resync_after_online_grow(mdev); + } else set_bit(RESYNC_AFTER_NEG, &mdev->flags); } } @@ -3118,13 +3235,16 @@ static int receive_state(struct drbd_conf *mdev, struct p_header *h) put_ldev(mdev); if (nconn == C_MASK) { + nconn = C_CONNECTED; if (mdev->state.disk == D_NEGOTIATING) { drbd_force_state(mdev, NS(disk, D_DISKLESS)); - nconn = C_CONNECTED; } else if (peer_state.disk == D_NEGOTIATING) { dev_err(DEV, "Disk attach process on the peer node was aborted.\n"); peer_state.disk = D_DISKLESS; + real_peer_disk = D_DISKLESS; } else { + if (test_and_clear_bit(CONN_DRY_RUN, &mdev->flags)) + return FALSE; D_ASSERT(oconn == C_WF_REPORT_PARAMS); drbd_force_state(mdev, NS(conn, C_DISCONNECTING)); return FALSE; @@ -3466,6 +3586,92 @@ static int receive_UnplugRemote(struct drbd_conf *mdev, struct p_header *h) return TRUE; } +static void timeval_sub_us(struct timeval* tv, unsigned int us) +{ + tv->tv_sec -= us / 1000000; + us = us % 1000000; + if (tv->tv_usec > us) { + tv->tv_usec += 1000000; + tv->tv_sec--; + } + tv->tv_usec -= us; +} + +static void got_delay_probe(struct drbd_conf *mdev, int from, struct p_delay_probe *p) +{ + struct delay_probe *dp; + struct list_head *le; + struct timeval now; + int seq_num; + int offset; + int data_delay; + + seq_num = be32_to_cpu(p->seq_num); + offset = be32_to_cpu(p->offset); + + spin_lock(&mdev->peer_seq_lock); + if (!list_empty(&mdev->delay_probes)) { + if (from == USE_DATA_SOCKET) + le = mdev->delay_probes.next; + else + le = mdev->delay_probes.prev; + + dp = list_entry(le, struct delay_probe, list); + + if (dp->seq_num == seq_num) { + list_del(le); + spin_unlock(&mdev->peer_seq_lock); + do_gettimeofday(&now); + timeval_sub_us(&now, offset); + data_delay = + now.tv_usec - dp->time.tv_usec + + (now.tv_sec - dp->time.tv_sec) * 1000000; + + if (data_delay > 0) + mdev->data_delay = data_delay; + + kfree(dp); + return; + } + + if (dp->seq_num > seq_num) { + spin_unlock(&mdev->peer_seq_lock); + dev_warn(DEV, "Previous allocation failure of struct delay_probe?\n"); + return; /* Do not alloca a struct delay_probe.... */ + } + } + spin_unlock(&mdev->peer_seq_lock); + + dp = kmalloc(sizeof(struct delay_probe), GFP_NOIO); + if (!dp) { + dev_warn(DEV, "Failed to allocate a struct delay_probe, do not worry.\n"); + return; + } + + dp->seq_num = seq_num; + do_gettimeofday(&dp->time); + timeval_sub_us(&dp->time, offset); + + spin_lock(&mdev->peer_seq_lock); + if (from == USE_DATA_SOCKET) + list_add(&dp->list, &mdev->delay_probes); + else + list_add_tail(&dp->list, &mdev->delay_probes); + spin_unlock(&mdev->peer_seq_lock); +} + +static int receive_delay_probe(struct drbd_conf *mdev, struct p_header *h) +{ + struct p_delay_probe *p = (struct p_delay_probe *)h; + + ERR_IF(h->length != (sizeof(*p)-sizeof(*h))) return FALSE; + if (drbd_recv(mdev, h->payload, h->length) != h->length) + return FALSE; + + got_delay_probe(mdev, USE_DATA_SOCKET, p); + return TRUE; +} + typedef int (*drbd_cmd_handler_f)(struct drbd_conf *, struct p_header *); static drbd_cmd_handler_f drbd_default_handler[] = { @@ -3489,6 +3695,7 @@ static drbd_cmd_handler_f drbd_default_handler[] = { [P_OV_REQUEST] = receive_DataRequest, [P_OV_REPLY] = receive_DataRequest, [P_CSUM_RS_REQUEST] = receive_DataRequest, + [P_DELAY_PROBE] = receive_delay_probe, /* anything missing from this table is in * the asender_tbl, see get_asender_cmd */ [P_MAX_CMD] = NULL, @@ -3594,10 +3801,7 @@ static void drbd_disconnect(struct drbd_conf *mdev) /* asender does not clean up anything. it must not interfere, either */ drbd_thread_stop(&mdev->asender); - - mutex_lock(&mdev->data.mutex); drbd_free_sock(mdev); - mutex_unlock(&mdev->data.mutex); spin_lock_irq(&mdev->req_lock); _drbd_wait_ee_list_empty(mdev, &mdev->active_ee); @@ -3718,7 +3922,7 @@ static void drbd_disconnect(struct drbd_conf *mdev) dev_info(DEV, "net_ee not empty, killed %u entries\n", i); i = atomic_read(&mdev->pp_in_use); if (i) - dev_info(DEV, "pp_in_use = %u, expected 0\n", i); + dev_info(DEV, "pp_in_use = %d, expected 0\n", i); D_ASSERT(list_empty(&mdev->read_ee)); D_ASSERT(list_empty(&mdev->active_ee)); @@ -4054,6 +4258,8 @@ static int got_PingAck(struct drbd_conf *mdev, struct p_header *h) { /* restore idle timeout */ mdev->meta.socket->sk->sk_rcvtimeo = mdev->net_conf->ping_int*HZ; + if (!test_and_set_bit(GOT_PING_ACK, &mdev->flags)) + wake_up(&mdev->misc_wait); return TRUE; } @@ -4209,7 +4415,6 @@ static int got_NegRSDReply(struct drbd_conf *mdev, struct p_header *h) sector = be64_to_cpu(p->sector); size = be32_to_cpu(p->blksize); - D_ASSERT(p->block_id == ID_SYNCER); update_peer_seq(mdev, be32_to_cpu(p->seq_num)); @@ -4267,6 +4472,14 @@ static int got_OVResult(struct drbd_conf *mdev, struct p_header *h) return TRUE; } +static int got_delay_probe_m(struct drbd_conf *mdev, struct p_header *h) +{ + struct p_delay_probe *p = (struct p_delay_probe *)h; + + got_delay_probe(mdev, USE_META_SOCKET, p); + return TRUE; +} + struct asender_cmd { size_t pkt_size; int (*process)(struct drbd_conf *mdev, struct p_header *h); @@ -4291,6 +4504,7 @@ static struct asender_cmd *get_asender_cmd(int cmd) [P_BARRIER_ACK] = { sizeof(struct p_barrier_ack), got_BarrierAck }, [P_STATE_CHG_REPLY] = { sizeof(struct p_req_state_reply), got_RqSReply }, [P_RS_IS_IN_SYNC] = { sizeof(struct p_block_ack), got_IsInSync }, + [P_DELAY_PROBE] = { sizeof(struct p_delay_probe), got_delay_probe_m }, [P_MAX_CMD] = { 0, NULL }, }; if (cmd > P_MAX_CMD || asender_tbl[cmd].process == NULL) diff --git a/drivers/block/drbd/drbd_req.c b/drivers/block/drbd/drbd_req.c index de81ab7b4627..654f1ef5cbb0 100644 --- a/drivers/block/drbd/drbd_req.c +++ b/drivers/block/drbd/drbd_req.c @@ -102,32 +102,7 @@ static void _req_is_done(struct drbd_conf *mdev, struct drbd_request *req, const } } - /* if it was a local io error, we want to notify our - * peer about that, and see if we need to - * detach the disk and stuff. - * to avoid allocating some special work - * struct, reuse the request. */ - - /* THINK - * why do we do this not when we detect the error, - * but delay it until it is "done", i.e. possibly - * until the next barrier ack? */ - - if (rw == WRITE && - ((s & RQ_LOCAL_MASK) && !(s & RQ_LOCAL_OK))) { - if (!(req->w.list.next == LIST_POISON1 || - list_empty(&req->w.list))) { - /* DEBUG ASSERT only; if this triggers, we - * probably corrupt the worker list here */ - dev_err(DEV, "req->w.list.next = %p\n", req->w.list.next); - dev_err(DEV, "req->w.list.prev = %p\n", req->w.list.prev); - } - req->w.cb = w_io_error; - drbd_queue_work(&mdev->data.work, &req->w); - /* drbd_req_free() is done in w_io_error */ - } else { - drbd_req_free(req); - } + drbd_req_free(req); } static void queue_barrier(struct drbd_conf *mdev) @@ -453,9 +428,6 @@ void __req_mod(struct drbd_request *req, enum drbd_req_event what, req->rq_state |= RQ_LOCAL_COMPLETED; req->rq_state &= ~RQ_LOCAL_PENDING; - dev_alert(DEV, "Local WRITE failed sec=%llus size=%u\n", - (unsigned long long)req->sector, req->size); - /* and now: check how to handle local io error. */ __drbd_chk_io_error(mdev, FALSE); _req_may_be_done(req, m); put_ldev(mdev); @@ -475,22 +447,21 @@ void __req_mod(struct drbd_request *req, enum drbd_req_event what, req->rq_state |= RQ_LOCAL_COMPLETED; req->rq_state &= ~RQ_LOCAL_PENDING; - dev_alert(DEV, "Local READ failed sec=%llus size=%u\n", - (unsigned long long)req->sector, req->size); - /* _req_mod(req,to_be_send); oops, recursion... */ D_ASSERT(!(req->rq_state & RQ_NET_MASK)); - req->rq_state |= RQ_NET_PENDING; - inc_ap_pending(mdev); __drbd_chk_io_error(mdev, FALSE); put_ldev(mdev); - /* NOTE: if we have no connection, - * or know the peer has no good data either, - * then we don't actually need to "queue_for_net_read", - * but we do so anyways, since the drbd_io_error() - * and the potential state change to "Diskless" - * needs to be done from process context */ + /* no point in retrying if there is no good remote data, + * or we have no connection. */ + if (mdev->state.pdsk != D_UP_TO_DATE) { + _req_may_be_done(req, m); + break; + } + + /* _req_mod(req,to_be_send); oops, recursion... */ + req->rq_state |= RQ_NET_PENDING; + inc_ap_pending(mdev); /* fall through: _req_mod(req,queue_for_net_read); */ case queue_for_net_read: @@ -600,6 +571,9 @@ void __req_mod(struct drbd_request *req, enum drbd_req_event what, _req_may_be_done(req, m); break; + case read_retry_remote_canceled: + req->rq_state &= ~RQ_NET_QUEUED; + /* fall through, in case we raced with drbd_disconnect */ case connection_lost_while_pending: /* transfer log cleanup after connection loss */ /* assert something? */ @@ -722,6 +696,7 @@ static int drbd_make_request_common(struct drbd_conf *mdev, struct bio *bio) struct drbd_request *req; int local, remote; int err = -EIO; + int ret = 0; /* allocate outside of all locks; */ req = drbd_req_new(mdev, bio); @@ -784,7 +759,7 @@ static int drbd_make_request_common(struct drbd_conf *mdev, struct bio *bio) (mdev->state.pdsk == D_INCONSISTENT && mdev->state.conn >= C_CONNECTED)); - if (!(local || remote)) { + if (!(local || remote) && !mdev->state.susp) { dev_err(DEV, "IO ERROR: neither local nor remote disk\n"); goto fail_free_complete; } @@ -810,6 +785,16 @@ allocate_barrier: /* GOOD, everything prepared, grab the spin_lock */ spin_lock_irq(&mdev->req_lock); + if (mdev->state.susp) { + /* If we got suspended, use the retry mechanism of + generic_make_request() to restart processing of this + bio. In the next call to drbd_make_request_26 + we sleep in inc_ap_bio() */ + ret = 1; + spin_unlock_irq(&mdev->req_lock); + goto fail_free_complete; + } + if (remote) { remote = (mdev->state.pdsk == D_UP_TO_DATE || (mdev->state.pdsk == D_INCONSISTENT && @@ -947,12 +932,14 @@ fail_and_free_req: req->private_bio = NULL; put_ldev(mdev); } - bio_endio(bio, err); + if (!ret) + bio_endio(bio, err); + drbd_req_free(req); dec_ap_bio(mdev); kfree(b); - return 0; + return ret; } /* helper function for drbd_make_request @@ -962,11 +949,6 @@ fail_and_free_req: */ static int drbd_fail_request_early(struct drbd_conf *mdev, int is_write) { - /* Unconfigured */ - if (mdev->state.conn == C_DISCONNECTING && - mdev->state.disk == D_DISKLESS) - return 1; - if (mdev->state.role != R_PRIMARY && (!allow_oos || is_write)) { if (__ratelimit(&drbd_ratelimit_state)) { @@ -1070,15 +1052,21 @@ int drbd_make_request_26(struct request_queue *q, struct bio *bio) /* we need to get a "reference count" (ap_bio_cnt) * to avoid races with the disconnect/reconnect/suspend code. - * In case we need to split the bio here, we need to get two references + * In case we need to split the bio here, we need to get three references * atomically, otherwise we might deadlock when trying to submit the * second one! */ - inc_ap_bio(mdev, 2); + inc_ap_bio(mdev, 3); D_ASSERT(e_enr == s_enr + 1); - drbd_make_request_common(mdev, &bp->bio1); - drbd_make_request_common(mdev, &bp->bio2); + while (drbd_make_request_common(mdev, &bp->bio1)) + inc_ap_bio(mdev, 1); + + while (drbd_make_request_common(mdev, &bp->bio2)) + inc_ap_bio(mdev, 1); + + dec_ap_bio(mdev); + bio_pair_release(bp); } return 0; @@ -1115,7 +1103,7 @@ int drbd_merge_bvec(struct request_queue *q, struct bvec_merge_data *bvm, struct } else if (limit && get_ldev(mdev)) { struct request_queue * const b = mdev->ldev->backing_bdev->bd_disk->queue; - if (b->merge_bvec_fn && mdev->ldev->dc.use_bmbv) { + if (b->merge_bvec_fn) { backing_limit = b->merge_bvec_fn(b, bvm, bvec); limit = min(limit, backing_limit); } diff --git a/drivers/block/drbd/drbd_req.h b/drivers/block/drbd/drbd_req.h index f22c1bc8ec7e..02d575d24518 100644 --- a/drivers/block/drbd/drbd_req.h +++ b/drivers/block/drbd/drbd_req.h @@ -57,7 +57,7 @@ * * It may me handed over to the local disk subsystem. * It may be completed by the local disk subsystem, - * either sucessfully or with io-error. + * either successfully or with io-error. * In case it is a READ request, and it failed locally, * it may be retried remotely. * @@ -91,6 +91,7 @@ enum drbd_req_event { send_failed, handed_over_to_network, connection_lost_while_pending, + read_retry_remote_canceled, recv_acked_by_peer, write_acked_by_peer, write_acked_by_peer_and_sis, /* and set_in_sync */ diff --git a/drivers/block/drbd/drbd_strings.c b/drivers/block/drbd/drbd_strings.c index 76863e3f05be..85179e1fb50a 100644 --- a/drivers/block/drbd/drbd_strings.c +++ b/drivers/block/drbd/drbd_strings.c @@ -70,7 +70,7 @@ static const char *drbd_disk_s_names[] = { static const char *drbd_state_sw_errors[] = { [-SS_TWO_PRIMARIES] = "Multiple primaries not allowed by config", - [-SS_NO_UP_TO_DATE_DISK] = "Refusing to be Primary without at least one UpToDate disk", + [-SS_NO_UP_TO_DATE_DISK] = "Need access to UpToDate data", [-SS_NO_LOCAL_DISK] = "Can not resync without local disk", [-SS_NO_REMOTE_DISK] = "Can not resync without remote disk", [-SS_CONNECTED_OUTDATES] = "Refusing to be Outdated while Connected", diff --git a/drivers/block/drbd/drbd_worker.c b/drivers/block/drbd/drbd_worker.c index b453c2bca3be..b623ceee2a4a 100644 --- a/drivers/block/drbd/drbd_worker.c +++ b/drivers/block/drbd/drbd_worker.c @@ -47,8 +47,7 @@ static int w_make_ov_request(struct drbd_conf *mdev, struct drbd_work *w, int ca /* defined here: drbd_md_io_complete - drbd_endio_write_sec - drbd_endio_read_sec + drbd_endio_sec drbd_endio_pri * more endio handlers: @@ -85,27 +84,10 @@ void drbd_md_io_complete(struct bio *bio, int error) /* reads on behalf of the partner, * "submitted" by the receiver */ -void drbd_endio_read_sec(struct bio *bio, int error) __releases(local) +void drbd_endio_read_sec_final(struct drbd_epoch_entry *e) __releases(local) { unsigned long flags = 0; - struct drbd_epoch_entry *e = NULL; - struct drbd_conf *mdev; - int uptodate = bio_flagged(bio, BIO_UPTODATE); - - e = bio->bi_private; - mdev = e->mdev; - - if (error) - dev_warn(DEV, "read: error=%d s=%llus\n", error, - (unsigned long long)e->sector); - if (!error && !uptodate) { - dev_warn(DEV, "read: setting error to -EIO s=%llus\n", - (unsigned long long)e->sector); - /* strange behavior of some lower level drivers... - * fail the request by clearing the uptodate flag, - * but do not return any error?! */ - error = -EIO; - } + struct drbd_conf *mdev = e->mdev; D_ASSERT(e->block_id != ID_VACANT); @@ -114,49 +96,38 @@ void drbd_endio_read_sec(struct bio *bio, int error) __releases(local) list_del(&e->w.list); if (list_empty(&mdev->read_ee)) wake_up(&mdev->ee_wait); + if (test_bit(__EE_WAS_ERROR, &e->flags)) + __drbd_chk_io_error(mdev, FALSE); spin_unlock_irqrestore(&mdev->req_lock, flags); - drbd_chk_io_error(mdev, error, FALSE); drbd_queue_work(&mdev->data.work, &e->w); put_ldev(mdev); } +static int is_failed_barrier(int ee_flags) +{ + return (ee_flags & (EE_IS_BARRIER|EE_WAS_ERROR|EE_RESUBMITTED)) + == (EE_IS_BARRIER|EE_WAS_ERROR); +} + /* writes on behalf of the partner, or resync writes, - * "submitted" by the receiver. - */ -void drbd_endio_write_sec(struct bio *bio, int error) __releases(local) + * "submitted" by the receiver, final stage. */ +static void drbd_endio_write_sec_final(struct drbd_epoch_entry *e) __releases(local) { unsigned long flags = 0; - struct drbd_epoch_entry *e = NULL; - struct drbd_conf *mdev; + struct drbd_conf *mdev = e->mdev; sector_t e_sector; int do_wake; int is_syncer_req; int do_al_complete_io; - int uptodate = bio_flagged(bio, BIO_UPTODATE); - int is_barrier = bio_rw_flagged(bio, BIO_RW_BARRIER); - - e = bio->bi_private; - mdev = e->mdev; - - if (error) - dev_warn(DEV, "write: error=%d s=%llus\n", error, - (unsigned long long)e->sector); - if (!error && !uptodate) { - dev_warn(DEV, "write: setting error to -EIO s=%llus\n", - (unsigned long long)e->sector); - /* strange behavior of some lower level drivers... - * fail the request by clearing the uptodate flag, - * but do not return any error?! */ - error = -EIO; - } - /* error == -ENOTSUPP would be a better test, - * alas it is not reliable */ - if (error && is_barrier && e->flags & EE_IS_BARRIER) { + /* if this is a failed barrier request, disable use of barriers, + * and schedule for resubmission */ + if (is_failed_barrier(e->flags)) { drbd_bump_write_ordering(mdev, WO_bdev_flush); spin_lock_irqsave(&mdev->req_lock, flags); list_del(&e->w.list); + e->flags = (e->flags & ~EE_WAS_ERROR) | EE_RESUBMITTED; e->w.cb = w_e_reissue; /* put_ldev actually happens below, once we come here again. */ __release(local); @@ -167,17 +138,16 @@ void drbd_endio_write_sec(struct bio *bio, int error) __releases(local) D_ASSERT(e->block_id != ID_VACANT); - spin_lock_irqsave(&mdev->req_lock, flags); - mdev->writ_cnt += e->size >> 9; - is_syncer_req = is_syncer_block_id(e->block_id); - /* after we moved e to done_ee, * we may no longer access it, * it may be freed/reused already! * (as soon as we release the req_lock) */ e_sector = e->sector; do_al_complete_io = e->flags & EE_CALL_AL_COMPLETE_IO; + is_syncer_req = is_syncer_block_id(e->block_id); + spin_lock_irqsave(&mdev->req_lock, flags); + mdev->writ_cnt += e->size >> 9; list_del(&e->w.list); /* has been on active_ee or sync_ee */ list_add_tail(&e->w.list, &mdev->done_ee); @@ -190,7 +160,7 @@ void drbd_endio_write_sec(struct bio *bio, int error) __releases(local) ? list_empty(&mdev->sync_ee) : list_empty(&mdev->active_ee); - if (error) + if (test_bit(__EE_WAS_ERROR, &e->flags)) __drbd_chk_io_error(mdev, FALSE); spin_unlock_irqrestore(&mdev->req_lock, flags); @@ -205,7 +175,42 @@ void drbd_endio_write_sec(struct bio *bio, int error) __releases(local) wake_asender(mdev); put_ldev(mdev); +} + +/* writes on behalf of the partner, or resync writes, + * "submitted" by the receiver. + */ +void drbd_endio_sec(struct bio *bio, int error) +{ + struct drbd_epoch_entry *e = bio->bi_private; + struct drbd_conf *mdev = e->mdev; + int uptodate = bio_flagged(bio, BIO_UPTODATE); + int is_write = bio_data_dir(bio) == WRITE; + + if (error) + dev_warn(DEV, "%s: error=%d s=%llus\n", + is_write ? "write" : "read", error, + (unsigned long long)e->sector); + if (!error && !uptodate) { + dev_warn(DEV, "%s: setting error to -EIO s=%llus\n", + is_write ? "write" : "read", + (unsigned long long)e->sector); + /* strange behavior of some lower level drivers... + * fail the request by clearing the uptodate flag, + * but do not return any error?! */ + error = -EIO; + } + + if (error) + set_bit(__EE_WAS_ERROR, &e->flags); + bio_put(bio); /* no need for the bio anymore */ + if (atomic_dec_and_test(&e->pending_bios)) { + if (is_write) + drbd_endio_write_sec_final(e); + else + drbd_endio_read_sec_final(e); + } } /* read, readA or write requests on R_PRIMARY coming from drbd_make_request @@ -219,9 +224,6 @@ void drbd_endio_pri(struct bio *bio, int error) enum drbd_req_event what; int uptodate = bio_flagged(bio, BIO_UPTODATE); - if (error) - dev_warn(DEV, "p %s: error=%d\n", - bio_data_dir(bio) == WRITE ? "write" : "read", error); if (!error && !uptodate) { dev_warn(DEV, "p %s: setting error to -EIO\n", bio_data_dir(bio) == WRITE ? "write" : "read"); @@ -235,7 +237,7 @@ void drbd_endio_pri(struct bio *bio, int error) if (unlikely(error)) { what = (bio_data_dir(bio) == WRITE) ? write_completed_with_error - : (bio_rw(bio) == READA) + : (bio_rw(bio) == READ) ? read_completed_with_error : read_ahead_completed_with_error; } else @@ -252,20 +254,6 @@ void drbd_endio_pri(struct bio *bio, int error) complete_master_bio(mdev, &m); } -int w_io_error(struct drbd_conf *mdev, struct drbd_work *w, int cancel) -{ - struct drbd_request *req = container_of(w, struct drbd_request, w); - - /* NOTE: mdev->ldev can be NULL by the time we get here! */ - /* D_ASSERT(mdev->ldev->dc.on_io_error != EP_PASS_ON); */ - - /* the only way this callback is scheduled is from _req_may_be_done, - * when it is done and had a local write error, see comments there */ - drbd_req_free(req); - - return TRUE; -} - int w_read_retry_remote(struct drbd_conf *mdev, struct drbd_work *w, int cancel) { struct drbd_request *req = container_of(w, struct drbd_request, w); @@ -275,12 +263,9 @@ int w_read_retry_remote(struct drbd_conf *mdev, struct drbd_work *w, int cancel) * to give the disk the chance to relocate that block */ spin_lock_irq(&mdev->req_lock); - if (cancel || - mdev->state.conn < C_CONNECTED || - mdev->state.pdsk <= D_INCONSISTENT) { - _req_mod(req, send_canceled); + if (cancel || mdev->state.pdsk != D_UP_TO_DATE) { + _req_mod(req, read_retry_remote_canceled); spin_unlock_irq(&mdev->req_lock); - dev_alert(DEV, "WE ARE LOST. Local IO failure, no peer.\n"); return 1; } spin_unlock_irq(&mdev->req_lock); @@ -295,7 +280,34 @@ int w_resync_inactive(struct drbd_conf *mdev, struct drbd_work *w, int cancel) return 1; /* Simply ignore this! */ } -void drbd_csum(struct drbd_conf *mdev, struct crypto_hash *tfm, struct bio *bio, void *digest) +void drbd_csum_ee(struct drbd_conf *mdev, struct crypto_hash *tfm, struct drbd_epoch_entry *e, void *digest) +{ + struct hash_desc desc; + struct scatterlist sg; + struct page *page = e->pages; + struct page *tmp; + unsigned len; + + desc.tfm = tfm; + desc.flags = 0; + + sg_init_table(&sg, 1); + crypto_hash_init(&desc); + + while ((tmp = page_chain_next(page))) { + /* all but the last page will be fully used */ + sg_set_page(&sg, page, PAGE_SIZE, 0); + crypto_hash_update(&desc, &sg, sg.length); + page = tmp; + } + /* and now the last, possibly only partially used page */ + len = e->size & (PAGE_SIZE - 1); + sg_set_page(&sg, page, len ?: PAGE_SIZE, 0); + crypto_hash_update(&desc, &sg, sg.length); + crypto_hash_final(&desc, digest); +} + +void drbd_csum_bio(struct drbd_conf *mdev, struct crypto_hash *tfm, struct bio *bio, void *digest) { struct hash_desc desc; struct scatterlist sg; @@ -329,11 +341,11 @@ static int w_e_send_csum(struct drbd_conf *mdev, struct drbd_work *w, int cancel return 1; } - if (likely(drbd_bio_uptodate(e->private_bio))) { + if (likely((e->flags & EE_WAS_ERROR) == 0)) { digest_size = crypto_hash_digestsize(mdev->csums_tfm); digest = kmalloc(digest_size, GFP_NOIO); if (digest) { - drbd_csum(mdev, mdev->csums_tfm, e->private_bio, digest); + drbd_csum_ee(mdev, mdev->csums_tfm, e, digest); inc_rs_pending(mdev); ok = drbd_send_drequest_csum(mdev, @@ -369,23 +381,21 @@ static int read_for_csum(struct drbd_conf *mdev, sector_t sector, int size) /* GFP_TRY, because if there is no memory available right now, this may * be rescheduled for later. It is "only" background resync, after all. */ e = drbd_alloc_ee(mdev, DRBD_MAGIC+0xbeef, sector, size, GFP_TRY); - if (!e) { - put_ldev(mdev); - return 2; - } + if (!e) + goto fail; spin_lock_irq(&mdev->req_lock); list_add(&e->w.list, &mdev->read_ee); spin_unlock_irq(&mdev->req_lock); - e->private_bio->bi_end_io = drbd_endio_read_sec; - e->private_bio->bi_rw = READ; e->w.cb = w_e_send_csum; + if (drbd_submit_ee(mdev, e, READ, DRBD_FAULT_RS_RD) == 0) + return 1; - mdev->read_cnt += size >> 9; - drbd_generic_make_request(mdev, DRBD_FAULT_RS_RD, e->private_bio); - - return 1; + drbd_free_ee(mdev, e); +fail: + put_ldev(mdev); + return 2; } void resync_timer_fn(unsigned long data) @@ -414,13 +424,25 @@ void resync_timer_fn(unsigned long data) drbd_queue_work(&mdev->data.work, &mdev->resync_work); } +static int calc_resync_rate(struct drbd_conf *mdev) +{ + int d = mdev->data_delay / 1000; /* us -> ms */ + int td = mdev->sync_conf.throttle_th * 100; /* 0.1s -> ms */ + int hd = mdev->sync_conf.hold_off_th * 100; /* 0.1s -> ms */ + int cr = mdev->sync_conf.rate; + + return d <= td ? cr : + d >= hd ? 0 : + cr + (cr * (td - d) / (hd - td)); +} + int w_make_resync_request(struct drbd_conf *mdev, struct drbd_work *w, int cancel) { unsigned long bit; sector_t sector; const sector_t capacity = drbd_get_capacity(mdev->this_bdev); - int max_segment_size = queue_max_segment_size(mdev->rq_queue); + int max_segment_size; int number, i, size, pe, mx; int align, queued, sndbuf; @@ -446,7 +468,13 @@ int w_make_resync_request(struct drbd_conf *mdev, return 1; } - number = SLEEP_TIME * mdev->sync_conf.rate / ((BM_BLOCK_SIZE/1024)*HZ); + /* starting with drbd 8.3.8, we can handle multi-bio EEs, + * if it should be necessary */ + max_segment_size = mdev->agreed_pro_version < 94 ? + queue_max_segment_size(mdev->rq_queue) : DRBD_MAX_SEGMENT_SIZE; + + mdev->c_sync_rate = calc_resync_rate(mdev); + number = SLEEP_TIME * mdev->c_sync_rate / ((BM_BLOCK_SIZE / 1024) * HZ); pe = atomic_read(&mdev->rs_pending_cnt); mutex_lock(&mdev->data.mutex); @@ -509,12 +537,6 @@ next_sector: * * Additionally always align bigger requests, in order to * be prepared for all stripe sizes of software RAIDs. - * - * we _do_ care about the agreed-upon q->max_segment_size - * here, as splitting up the requests on the other side is more - * difficult. the consequence is, that on lvm and md and other - * "indirect" devices, this is dead code, since - * q->max_segment_size will be PAGE_SIZE. */ align = 1; for (;;) { @@ -806,7 +828,7 @@ out: /* helper */ static void move_to_net_ee_or_free(struct drbd_conf *mdev, struct drbd_epoch_entry *e) { - if (drbd_bio_has_active_page(e->private_bio)) { + if (drbd_ee_has_active_page(e)) { /* This might happen if sendpage() has not finished */ spin_lock_irq(&mdev->req_lock); list_add_tail(&e->w.list, &mdev->net_ee); @@ -832,7 +854,7 @@ int w_e_end_data_req(struct drbd_conf *mdev, struct drbd_work *w, int cancel) return 1; } - if (likely(drbd_bio_uptodate(e->private_bio))) { + if (likely((e->flags & EE_WAS_ERROR) == 0)) { ok = drbd_send_block(mdev, P_DATA_REPLY, e); } else { if (__ratelimit(&drbd_ratelimit_state)) @@ -873,7 +895,7 @@ int w_e_end_rsdata_req(struct drbd_conf *mdev, struct drbd_work *w, int cancel) put_ldev(mdev); } - if (likely(drbd_bio_uptodate(e->private_bio))) { + if (likely((e->flags & EE_WAS_ERROR) == 0)) { if (likely(mdev->state.pdsk >= D_INCONSISTENT)) { inc_rs_pending(mdev); ok = drbd_send_block(mdev, P_RS_DATA_REPLY, e); @@ -921,7 +943,7 @@ int w_e_end_csum_rs_req(struct drbd_conf *mdev, struct drbd_work *w, int cancel) di = (struct digest_info *)(unsigned long)e->block_id; - if (likely(drbd_bio_uptodate(e->private_bio))) { + if (likely((e->flags & EE_WAS_ERROR) == 0)) { /* quick hack to try to avoid a race against reconfiguration. * a real fix would be much more involved, * introducing more locking mechanisms */ @@ -931,14 +953,15 @@ int w_e_end_csum_rs_req(struct drbd_conf *mdev, struct drbd_work *w, int cancel) digest = kmalloc(digest_size, GFP_NOIO); } if (digest) { - drbd_csum(mdev, mdev->csums_tfm, e->private_bio, digest); + drbd_csum_ee(mdev, mdev->csums_tfm, e, digest); eq = !memcmp(digest, di->digest, digest_size); kfree(digest); } if (eq) { drbd_set_in_sync(mdev, e->sector, e->size); - mdev->rs_same_csum++; + /* rs_same_csums unit is BM_BLOCK_SIZE */ + mdev->rs_same_csum += e->size >> BM_BLOCK_SHIFT; ok = drbd_send_ack(mdev, P_RS_IS_IN_SYNC, e); } else { inc_rs_pending(mdev); @@ -972,14 +995,14 @@ int w_e_end_ov_req(struct drbd_conf *mdev, struct drbd_work *w, int cancel) if (unlikely(cancel)) goto out; - if (unlikely(!drbd_bio_uptodate(e->private_bio))) + if (unlikely((e->flags & EE_WAS_ERROR) != 0)) goto out; digest_size = crypto_hash_digestsize(mdev->verify_tfm); /* FIXME if this allocation fails, online verify will not terminate! */ digest = kmalloc(digest_size, GFP_NOIO); if (digest) { - drbd_csum(mdev, mdev->verify_tfm, e->private_bio, digest); + drbd_csum_ee(mdev, mdev->verify_tfm, e, digest); inc_rs_pending(mdev); ok = drbd_send_drequest_csum(mdev, e->sector, e->size, digest, digest_size, P_OV_REPLY); @@ -1028,11 +1051,11 @@ int w_e_end_ov_reply(struct drbd_conf *mdev, struct drbd_work *w, int cancel) di = (struct digest_info *)(unsigned long)e->block_id; - if (likely(drbd_bio_uptodate(e->private_bio))) { + if (likely((e->flags & EE_WAS_ERROR) == 0)) { digest_size = crypto_hash_digestsize(mdev->verify_tfm); digest = kmalloc(digest_size, GFP_NOIO); if (digest) { - drbd_csum(mdev, mdev->verify_tfm, e->private_bio, digest); + drbd_csum_ee(mdev, mdev->verify_tfm, e, digest); D_ASSERT(digest_size == di->digest_size); eq = !memcmp(digest, di->digest, digest_size); @@ -1288,6 +1311,14 @@ int drbd_alter_sa(struct drbd_conf *mdev, int na) return retcode; } +static void ping_peer(struct drbd_conf *mdev) +{ + clear_bit(GOT_PING_ACK, &mdev->flags); + request_ping(mdev); + wait_event(mdev->misc_wait, + test_bit(GOT_PING_ACK, &mdev->flags) || mdev->state.conn < C_CONNECTED); +} + /** * drbd_start_resync() - Start the resync process * @mdev: DRBD device. @@ -1371,7 +1402,6 @@ void drbd_start_resync(struct drbd_conf *mdev, enum drbd_conns side) _drbd_pause_after(mdev); } write_unlock_irq(&global_state_lock); - drbd_state_unlock(mdev); put_ldev(mdev); if (r == SS_SUCCESS) { @@ -1382,11 +1412,8 @@ void drbd_start_resync(struct drbd_conf *mdev, enum drbd_conns side) if (mdev->rs_total == 0) { /* Peer still reachable? Beware of failing before-resync-target handlers! */ - request_ping(mdev); - __set_current_state(TASK_INTERRUPTIBLE); - schedule_timeout(mdev->net_conf->ping_timeo*HZ/9); /* 9 instead 10 */ + ping_peer(mdev); drbd_resync_finished(mdev); - return; } /* ns.conn may already be != mdev->state.conn, @@ -1398,6 +1425,7 @@ void drbd_start_resync(struct drbd_conf *mdev, enum drbd_conns side) drbd_md_sync(mdev); } + drbd_state_unlock(mdev); } int drbd_worker(struct drbd_thread *thi) diff --git a/drivers/block/drbd/drbd_wrappers.h b/drivers/block/drbd/drbd_wrappers.h index f93fa111ce50..defdb5013ea3 100644 --- a/drivers/block/drbd/drbd_wrappers.h +++ b/drivers/block/drbd/drbd_wrappers.h @@ -18,23 +18,9 @@ static inline void drbd_set_my_capacity(struct drbd_conf *mdev, #define drbd_bio_uptodate(bio) bio_flagged(bio, BIO_UPTODATE) -static inline int drbd_bio_has_active_page(struct bio *bio) -{ - struct bio_vec *bvec; - int i; - - __bio_for_each_segment(bvec, bio, i, 0) { - if (page_count(bvec->bv_page) > 1) - return 1; - } - - return 0; -} - /* bi_end_io handlers */ extern void drbd_md_io_complete(struct bio *bio, int error); -extern void drbd_endio_read_sec(struct bio *bio, int error); -extern void drbd_endio_write_sec(struct bio *bio, int error); +extern void drbd_endio_sec(struct bio *bio, int error); extern void drbd_endio_pri(struct bio *bio, int error); /* diff --git a/drivers/block/floppy.c b/drivers/block/floppy.c index 3266b4f65daa..90c4038702da 100644 --- a/drivers/block/floppy.c +++ b/drivers/block/floppy.c @@ -144,13 +144,23 @@ * Better audit of register_blkdev. */ -#define FLOPPY_SANITY_CHECK #undef FLOPPY_SILENT_DCL_CLEAR #define REALLY_SLOW_IO #define DEBUGT 2 -#define DCL_DEBUG /* debug disk change line */ + +#define DPRINT(format, args...) \ + pr_info("floppy%d: " format, current_drive, ##args) + +#define DCL_DEBUG /* debug disk change line */ +#ifdef DCL_DEBUG +#define debug_dcl(test, fmt, args...) \ + do { if ((test) & FD_DEBUG) DPRINT(fmt, ##args); } while (0) +#else +#define debug_dcl(test, fmt, args...) \ + do { if (0) DPRINT(fmt, ##args); } while (0) +#endif /* do print messages for unexpected interrupts */ static int print_unex = 1; @@ -180,6 +190,8 @@ static int print_unex = 1; #include <linux/mod_devicetable.h> #include <linux/buffer_head.h> /* for invalidate_buffers() */ #include <linux/mutex.h> +#include <linux/io.h> +#include <linux/uaccess.h> /* * PS/2 floppies have much slower step rates than regular floppies. @@ -191,8 +203,6 @@ static int slow_floppy; #include <asm/dma.h> #include <asm/irq.h> #include <asm/system.h> -#include <asm/io.h> -#include <asm/uaccess.h> static int FLOPPY_IRQ = 6; static int FLOPPY_DMA = 2; @@ -241,8 +251,6 @@ static int allowed_drive_mask = 0x33; static int irqdma_allocated; -#define DEVICE_NAME "floppy" - #include <linux/blkdev.h> #include <linux/blkpg.h> #include <linux/cdrom.h> /* for the compatibility eject ioctl */ @@ -250,7 +258,7 @@ static int irqdma_allocated; static struct request *current_req; static struct request_queue *floppy_queue; -static void do_fd_request(struct request_queue * q); +static void do_fd_request(struct request_queue *q); #ifndef fd_get_dma_residue #define fd_get_dma_residue() get_dma_residue(FLOPPY_DMA) @@ -263,7 +271,7 @@ static void do_fd_request(struct request_queue * q); #endif #ifndef fd_dma_mem_alloc -#define fd_dma_mem_alloc(size) __get_dma_pages(GFP_KERNEL,get_order(size)) +#define fd_dma_mem_alloc(size) __get_dma_pages(GFP_KERNEL, get_order(size)) #endif static inline void fallback_on_nodma_alloc(char **addr, size_t l) @@ -273,7 +281,7 @@ static inline void fallback_on_nodma_alloc(char **addr, size_t l) return; /* we have the memory */ if (can_use_virtual_dma != 2) return; /* no fallback allowed */ - printk("DMA memory shortage. Temporarily falling back on virtual DMA\n"); + pr_info("DMA memory shortage. Temporarily falling back on virtual DMA\n"); *addr = (char *)nodma_mem_alloc(l); #else return; @@ -283,59 +291,50 @@ static inline void fallback_on_nodma_alloc(char **addr, size_t l) /* End dma memory related stuff */ static unsigned long fake_change; -static int initialising = 1; +static bool initialized; -#define ITYPE(x) (((x)>>2) & 0x1f) -#define TOMINOR(x) ((x & 3) | ((x & 4) << 5)) -#define UNIT(x) ((x) & 0x03) /* drive on fdc */ -#define FDC(x) (((x) & 0x04) >> 2) /* fdc of drive */ +#define ITYPE(x) (((x) >> 2) & 0x1f) +#define TOMINOR(x) ((x & 3) | ((x & 4) << 5)) +#define UNIT(x) ((x) & 0x03) /* drive on fdc */ +#define FDC(x) (((x) & 0x04) >> 2) /* fdc of drive */ /* reverse mapping from unit and fdc to drive */ #define REVDRIVE(fdc, unit) ((unit) + ((fdc) << 2)) -#define DP (&drive_params[current_drive]) -#define DRS (&drive_state[current_drive]) -#define DRWE (&write_errors[current_drive]) -#define FDCS (&fdc_state[fdc]) -#define CLEARF(x) clear_bit(x##_BIT, &DRS->flags) -#define SETF(x) set_bit(x##_BIT, &DRS->flags) -#define TESTF(x) test_bit(x##_BIT, &DRS->flags) -#define UDP (&drive_params[drive]) -#define UDRS (&drive_state[drive]) -#define UDRWE (&write_errors[drive]) -#define UFDCS (&fdc_state[FDC(drive)]) -#define UCLEARF(x) clear_bit(x##_BIT, &UDRS->flags) -#define USETF(x) set_bit(x##_BIT, &UDRS->flags) -#define UTESTF(x) test_bit(x##_BIT, &UDRS->flags) +#define DP (&drive_params[current_drive]) +#define DRS (&drive_state[current_drive]) +#define DRWE (&write_errors[current_drive]) +#define FDCS (&fdc_state[fdc]) -#define DPRINT(format, args...) printk(DEVICE_NAME "%d: " format, current_drive , ## args) +#define UDP (&drive_params[drive]) +#define UDRS (&drive_state[drive]) +#define UDRWE (&write_errors[drive]) +#define UFDCS (&fdc_state[FDC(drive)]) -#define PH_HEAD(floppy,head) (((((floppy)->stretch & 2) >>1) ^ head) << 2) -#define STRETCH(floppy) ((floppy)->stretch & FD_STRETCH) - -#define CLEARSTRUCT(x) memset((x), 0, sizeof(*(x))) +#define PH_HEAD(floppy, head) (((((floppy)->stretch & 2) >> 1) ^ head) << 2) +#define STRETCH(floppy) ((floppy)->stretch & FD_STRETCH) /* read/write */ -#define COMMAND raw_cmd->cmd[0] -#define DR_SELECT raw_cmd->cmd[1] -#define TRACK raw_cmd->cmd[2] -#define HEAD raw_cmd->cmd[3] -#define SECTOR raw_cmd->cmd[4] -#define SIZECODE raw_cmd->cmd[5] -#define SECT_PER_TRACK raw_cmd->cmd[6] -#define GAP raw_cmd->cmd[7] -#define SIZECODE2 raw_cmd->cmd[8] +#define COMMAND (raw_cmd->cmd[0]) +#define DR_SELECT (raw_cmd->cmd[1]) +#define TRACK (raw_cmd->cmd[2]) +#define HEAD (raw_cmd->cmd[3]) +#define SECTOR (raw_cmd->cmd[4]) +#define SIZECODE (raw_cmd->cmd[5]) +#define SECT_PER_TRACK (raw_cmd->cmd[6]) +#define GAP (raw_cmd->cmd[7]) +#define SIZECODE2 (raw_cmd->cmd[8]) #define NR_RW 9 /* format */ -#define F_SIZECODE raw_cmd->cmd[2] -#define F_SECT_PER_TRACK raw_cmd->cmd[3] -#define F_GAP raw_cmd->cmd[4] -#define F_FILL raw_cmd->cmd[5] +#define F_SIZECODE (raw_cmd->cmd[2]) +#define F_SECT_PER_TRACK (raw_cmd->cmd[3]) +#define F_GAP (raw_cmd->cmd[4]) +#define F_FILL (raw_cmd->cmd[5]) #define NR_F 6 /* - * Maximum disk size (in kilobytes). This default is used whenever the - * current disk size is unknown. + * Maximum disk size (in kilobytes). + * This default is used whenever the current disk size is unknown. * [Now it is rather a minimum] */ #define MAX_DISK_SIZE 4 /* 3984 */ @@ -345,16 +344,17 @@ static int initialising = 1; */ #define MAX_REPLIES 16 static unsigned char reply_buffer[MAX_REPLIES]; -static int inr; /* size of reply buffer, when called from interrupt */ -#define ST0 (reply_buffer[0]) -#define ST1 (reply_buffer[1]) -#define ST2 (reply_buffer[2]) -#define ST3 (reply_buffer[0]) /* result of GETSTATUS */ -#define R_TRACK (reply_buffer[3]) -#define R_HEAD (reply_buffer[4]) -#define R_SECTOR (reply_buffer[5]) -#define R_SIZECODE (reply_buffer[6]) -#define SEL_DLY (2*HZ/100) +static int inr; /* size of reply buffer, when called from interrupt */ +#define ST0 (reply_buffer[0]) +#define ST1 (reply_buffer[1]) +#define ST2 (reply_buffer[2]) +#define ST3 (reply_buffer[0]) /* result of GETSTATUS */ +#define R_TRACK (reply_buffer[3]) +#define R_HEAD (reply_buffer[4]) +#define R_SECTOR (reply_buffer[5]) +#define R_SIZECODE (reply_buffer[6]) + +#define SEL_DLY (2 * HZ / 100) /* * this struct defines the different floppy drive types. @@ -505,9 +505,9 @@ static char floppy_device_name[] = "floppy"; static int probing; /* Synchronization of FDC access. */ -#define FD_COMMAND_NONE -1 -#define FD_COMMAND_ERROR 2 -#define FD_COMMAND_OKAY 3 +#define FD_COMMAND_NONE -1 +#define FD_COMMAND_ERROR 2 +#define FD_COMMAND_OKAY 3 static volatile int command_status = FD_COMMAND_NONE; static unsigned long fdc_busy; @@ -515,11 +515,6 @@ static DECLARE_WAIT_QUEUE_HEAD(fdc_wait); static DECLARE_WAIT_QUEUE_HEAD(command_done); #define NO_SIGNAL (!interruptible || !signal_pending(current)) -#define CALL(x) if ((x) == -EINTR) return -EINTR -#define ECALL(x) if ((ret = (x))) return ret; -#define _WAIT(x,i) CALL(ret=wait_til_done((x),i)) -#define WAIT(x) _WAIT((x),interruptible) -#define IWAIT(x) _WAIT((x),1) /* Errors during formatting are counted here. */ static int format_errors; @@ -545,8 +540,9 @@ static int max_buffer_sectors; static int *errors; typedef void (*done_f)(int); static struct cont_t { - void (*interrupt)(void); /* this is called after the interrupt of the - * main command */ + void (*interrupt)(void); + /* this is called after the interrupt of the + * main command */ void (*redo)(void); /* this is called to retry the operation */ void (*error)(void); /* this is called to tally an error */ done_f done; /* this is called to say if the operation has @@ -571,7 +567,6 @@ static void floppy_release_irq_and_dma(void); * reset doesn't need to be tested before sending commands, because * output_byte is automatically disabled when reset is set. */ -#define CHECK_RESET { if (FDCS->reset){ reset_fdc(); return; } } static void reset_fdc(void); /* @@ -579,9 +574,9 @@ static void reset_fdc(void); * information to interrupts. They are the data used for the current * request. */ -#define NO_TRACK -1 -#define NEED_1_RECAL -2 -#define NEED_2_RECAL -3 +#define NO_TRACK -1 +#define NEED_1_RECAL -2 +#define NEED_2_RECAL -3 static int usage_count; @@ -621,39 +616,35 @@ static inline void set_debugt(void) debugtimer = jiffies; } -static inline void debugt(const char *message) +static inline void debugt(const char *func, const char *msg) { if (DP->flags & DEBUGT) - printk("%s dtime=%lu\n", message, jiffies - debugtimer); + pr_info("%s:%s dtime=%lu\n", func, msg, jiffies - debugtimer); } #else static inline void set_debugt(void) { } -static inline void debugt(const char *message) { } +static inline void debugt(const char *func, const char *msg) { } #endif /* DEBUGT */ -typedef void (*timeout_fn) (unsigned long); +typedef void (*timeout_fn)(unsigned long); static DEFINE_TIMER(fd_timeout, floppy_shutdown, 0, 0); static const char *timeout_message; -#ifdef FLOPPY_SANITY_CHECK -static void is_alive(const char *message) +static void is_alive(const char *func, const char *message) { /* this routine checks whether the floppy driver is "alive" */ - if (test_bit(0, &fdc_busy) && command_status < 2 - && !timer_pending(&fd_timeout)) { - DPRINT("timeout handler died: %s\n", message); + if (test_bit(0, &fdc_busy) && command_status < 2 && + !timer_pending(&fd_timeout)) { + DPRINT("%s: timeout handler died. %s\n", func, message); } } -#endif -static void (*do_floppy) (void) = NULL; - -#ifdef FLOPPY_SANITY_CHECK +static void (*do_floppy)(void) = NULL; #define OLOGSIZE 20 -static void (*lasthandler) (void); +static void (*lasthandler)(void); static unsigned long interruptjiffies; static unsigned long resultjiffies; static int resultsize; @@ -666,12 +657,11 @@ static struct output_log { } output_log[OLOGSIZE]; static int output_log_pos; -#endif #define current_reqD -1 #define MAXTIMEOUT -2 -static void __reschedule_timeout(int drive, const char *message, int marg) +static void __reschedule_timeout(int drive, const char *message) { if (drive == current_reqD) drive = current_drive; @@ -682,25 +672,22 @@ static void __reschedule_timeout(int drive, const char *message, int marg) } else fd_timeout.expires = jiffies + UDP->timeout; add_timer(&fd_timeout); - if (UDP->flags & FD_DEBUG) { - DPRINT("reschedule timeout "); - printk(message, marg); - printk("\n"); - } + if (UDP->flags & FD_DEBUG) + DPRINT("reschedule timeout %s\n", message); timeout_message = message; } -static void reschedule_timeout(int drive, const char *message, int marg) +static void reschedule_timeout(int drive, const char *message) { unsigned long flags; spin_lock_irqsave(&floppy_lock, flags); - __reschedule_timeout(drive, message, marg); + __reschedule_timeout(drive, message); spin_unlock_irqrestore(&floppy_lock, flags); } -#define INFBOUND(a,b) (a)=max_t(int, a, b) -#define SUPBOUND(a,b) (a)=min_t(int, a, b) +#define INFBOUND(a, b) (a) = max_t(int, a, b) +#define SUPBOUND(a, b) (a) = min_t(int, a, b) /* * Bottom half floppy driver. @@ -739,7 +726,6 @@ static int disk_change(int drive) { int fdc = FDC(drive); -#ifdef FLOPPY_SANITY_CHECK if (time_before(jiffies, UDRS->select_date + UDP->select_delay)) DPRINT("WARNING disk change called early\n"); if (!(FDCS->dor & (0x10 << UNIT(drive))) || @@ -748,31 +734,27 @@ static int disk_change(int drive) DPRINT("drive=%d fdc=%d dor=%x\n", drive, FDC(drive), (unsigned int)FDCS->dor); } -#endif -#ifdef DCL_DEBUG - if (UDP->flags & FD_DEBUG) { - DPRINT("checking disk change line for drive %d\n", drive); - DPRINT("jiffies=%lu\n", jiffies); - DPRINT("disk change line=%x\n", fd_inb(FD_DIR) & 0x80); - DPRINT("flags=%lx\n", UDRS->flags); - } -#endif + debug_dcl(UDP->flags, + "checking disk change line for drive %d\n", drive); + debug_dcl(UDP->flags, "jiffies=%lu\n", jiffies); + debug_dcl(UDP->flags, "disk change line=%x\n", fd_inb(FD_DIR) & 0x80); + debug_dcl(UDP->flags, "flags=%lx\n", UDRS->flags); + if (UDP->flags & FD_BROKEN_DCL) - return UTESTF(FD_DISK_CHANGED); + return test_bit(FD_DISK_CHANGED_BIT, &UDRS->flags); if ((fd_inb(FD_DIR) ^ UDP->flags) & 0x80) { - USETF(FD_VERIFY); /* verify write protection */ - if (UDRS->maxblock) { - /* mark it changed */ - USETF(FD_DISK_CHANGED); - } + set_bit(FD_VERIFY_BIT, &UDRS->flags); + /* verify write protection */ + + if (UDRS->maxblock) /* mark it changed */ + set_bit(FD_DISK_CHANGED_BIT, &UDRS->flags); /* invalidate its geometry */ if (UDRS->keep_data >= 0) { if ((UDP->flags & FTD_MSG) && current_type[drive] != NULL) - DPRINT("Disk type is undefined after " - "disk change\n"); + DPRINT("Disk type is undefined after disk change\n"); current_type[drive] = NULL; floppy_sizes[TOMINOR(drive)] = MAX_DISK_SIZE << 1; } @@ -780,7 +762,7 @@ static int disk_change(int drive) return 1; } else { UDRS->last_checked = jiffies; - UCLEARF(FD_DISK_NEWCHANGE); + clear_bit(FD_DISK_NEWCHANGE_BIT, &UDRS->flags); } return 0; } @@ -790,6 +772,12 @@ static inline int is_selected(int dor, int unit) return ((dor & (0x10 << unit)) && (dor & 3) == unit); } +static bool is_ready_state(int status) +{ + int state = status & (STATUS_READY | STATUS_DIR | STATUS_DMA); + return state == STATUS_READY; +} + static int set_dor(int fdc, char mask, char data) { unsigned char unit; @@ -806,11 +794,8 @@ static int set_dor(int fdc, char mask, char data) unit = olddor & 0x3; if (is_selected(olddor, unit) && !is_selected(newdor, unit)) { drive = REVDRIVE(fdc, unit); -#ifdef DCL_DEBUG - if (UDP->flags & FD_DEBUG) { - DPRINT("calling disk change from set_dor\n"); - } -#endif + debug_dcl(UDP->flags, + "calling disk change from set_dor\n"); disk_change(drive); } FDCS->dor = newdor; @@ -834,8 +819,10 @@ static void twaddle(void) DRS->select_date = jiffies; } -/* reset all driver information about the current fdc. This is needed after - * a reset, and after a raw command. */ +/* + * Reset all driver information about the current fdc. + * This is needed after a reset, and after a raw command. + */ static void reset_fdc_info(int mode) { int drive; @@ -857,7 +844,7 @@ static void set_fdc(int drive) current_drive = drive; } if (fdc != 1 && fdc != 0) { - printk("bad fdc value\n"); + pr_info("bad fdc value\n"); return; } set_dor(fdc, ~0, 8); @@ -871,11 +858,10 @@ static void set_fdc(int drive) } /* locks the driver */ -static int _lock_fdc(int drive, int interruptible, int line) +static int _lock_fdc(int drive, bool interruptible, int line) { if (!usage_count) { - printk(KERN_ERR - "Trying to lock fdc while usage count=0 at line %d\n", + pr_err("Trying to lock fdc while usage count=0 at line %d\n", line); return -1; } @@ -904,15 +890,13 @@ static int _lock_fdc(int drive, int interruptible, int line) } command_status = FD_COMMAND_NONE; - __reschedule_timeout(drive, "lock fdc", 0); + __reschedule_timeout(drive, "lock fdc"); set_fdc(drive); return 0; } -#define lock_fdc(drive,interruptible) _lock_fdc(drive,interruptible, __LINE__) - -#define LOCK_FDC(drive,interruptible) \ -if (lock_fdc(drive,interruptible)) return -EINTR; +#define lock_fdc(drive, interruptible) \ + _lock_fdc(drive, interruptible, __LINE__) /* unlocks the driver */ static inline void unlock_fdc(void) @@ -924,7 +908,7 @@ static inline void unlock_fdc(void) DPRINT("FDC access conflict!\n"); if (do_floppy) - DPRINT("device interrupt still active at FDC release: %p!\n", + DPRINT("device interrupt still active at FDC release: %pf!\n", do_floppy); command_status = FD_COMMAND_NONE; spin_lock_irqsave(&floppy_lock, flags); @@ -1003,7 +987,7 @@ static void empty(void) static DECLARE_WORK(floppy_work, NULL); -static void schedule_bh(void (*handler) (void)) +static void schedule_bh(void (*handler)(void)) { PREPARE_WORK(&floppy_work, (work_func_t)handler); schedule_work(&floppy_work); @@ -1026,11 +1010,7 @@ static void cancel_activity(void) * transfer */ static void fd_watchdog(void) { -#ifdef DCL_DEBUG - if (DP->flags & FD_DEBUG) { - DPRINT("calling disk change from watchdog\n"); - } -#endif + debug_dcl(DP->flags, "calling disk change from watchdog\n"); if (disk_change(current_drive)) { DPRINT("disk removed during i/o\n"); @@ -1039,7 +1019,7 @@ static void fd_watchdog(void) reset_fdc(); } else { del_timer(&fd_timer); - fd_timer.function = (timeout_fn) fd_watchdog; + fd_timer.function = (timeout_fn)fd_watchdog; fd_timer.expires = jiffies + HZ / 10; add_timer(&fd_timer); } @@ -1105,25 +1085,23 @@ static void setup_DMA(void) { unsigned long f; -#ifdef FLOPPY_SANITY_CHECK if (raw_cmd->length == 0) { int i; - printk("zero dma transfer size:"); + pr_info("zero dma transfer size:"); for (i = 0; i < raw_cmd->cmd_count; i++) - printk("%x,", raw_cmd->cmd[i]); - printk("\n"); + pr_cont("%x,", raw_cmd->cmd[i]); + pr_cont("\n"); cont->done(0); FDCS->reset = 1; return; } if (((unsigned long)raw_cmd->kernel_data) % 512) { - printk("non aligned address: %p\n", raw_cmd->kernel_data); + pr_info("non aligned address: %p\n", raw_cmd->kernel_data); cont->done(0); FDCS->reset = 1; return; } -#endif f = claim_dma_lock(); fd_disable_dma(); #ifdef fd_dma_setup @@ -1165,7 +1143,7 @@ static int wait_til_ready(void) if (status & STATUS_READY) return status; } - if (!initialising) { + if (initialized) { DPRINT("Getstatus times out (%x) on fdc %d\n", status, fdc); show_floppy(); } @@ -1176,22 +1154,21 @@ static int wait_til_ready(void) /* sends a command byte to the fdc */ static int output_byte(char byte) { - int status; + int status = wait_til_ready(); - if ((status = wait_til_ready()) < 0) + if (status < 0) return -1; - if ((status & (STATUS_READY | STATUS_DIR | STATUS_DMA)) == STATUS_READY) { + + if (is_ready_state(status)) { fd_outb(byte, FD_DATA); -#ifdef FLOPPY_SANITY_CHECK output_log[output_log_pos].data = byte; output_log[output_log_pos].status = status; output_log[output_log_pos].jiffies = jiffies; output_log_pos = (output_log_pos + 1) % OLOGSIZE; -#endif return 0; } FDCS->reset = 1; - if (!initialising) { + if (initialized) { DPRINT("Unable to send byte %x to FDC. Fdc=%x Status=%x\n", byte, fdc, status); show_floppy(); @@ -1199,8 +1176,6 @@ static int output_byte(char byte) return -1; } -#define LAST_OUT(x) if (output_byte(x)<0){ reset_fdc();return;} - /* gets the response from the fdc */ static int result(void) { @@ -1208,14 +1183,13 @@ static int result(void) int status = 0; for (i = 0; i < MAX_REPLIES; i++) { - if ((status = wait_til_ready()) < 0) + status = wait_til_ready(); + if (status < 0) break; status &= STATUS_DIR | STATUS_READY | STATUS_BUSY | STATUS_DMA; if ((status & ~STATUS_BUSY) == STATUS_READY) { -#ifdef FLOPPY_SANITY_CHECK resultjiffies = jiffies; resultsize = i; -#endif return i; } if (status == (STATUS_DIR | STATUS_READY | STATUS_BUSY)) @@ -1223,10 +1197,9 @@ static int result(void) else break; } - if (!initialising) { - DPRINT - ("get result error. Fdc=%d Last status=%x Read bytes=%d\n", - fdc, status, i); + if (initialized) { + DPRINT("get result error. Fdc=%d Last status=%x Read bytes=%d\n", + fdc, status, i); show_floppy(); } FDCS->reset = 1; @@ -1237,12 +1210,14 @@ static int result(void) /* does the fdc need more output? */ static int need_more_output(void) { - int status; + int status = wait_til_ready(); - if ((status = wait_til_ready()) < 0) + if (status < 0) return -1; - if ((status & (STATUS_READY | STATUS_DIR | STATUS_DMA)) == STATUS_READY) + + if (is_ready_state(status)) return MORE_OUTPUT; + return result(); } @@ -1264,9 +1239,12 @@ static inline void perpendicular_mode(void) default: DPRINT("Invalid data rate for perpendicular mode!\n"); cont->done(0); - FDCS->reset = 1; /* convenient way to return to - * redo without to much hassle (deep - * stack et al. */ + FDCS->reset = 1; + /* + * convenient way to return to + * redo without too much hassle + * (deep stack et al.) + */ return; } } else @@ -1366,9 +1344,9 @@ static void fdc_specify(void) /* Convert step rate from microseconds to milliseconds and 4 bits */ srt = 16 - DIV_ROUND_UP(DP->srt * scale_dtr / 1000, NOMINAL_DTR); - if (slow_floppy) { + if (slow_floppy) srt = srt / 4; - } + SUPBOUND(srt, 0xf); INFBOUND(srt, 0); @@ -1415,16 +1393,46 @@ static int fdc_dtr(void) * Pause 5 msec to avoid trouble. (Needs to be 2 jiffies) */ FDCS->dtr = raw_cmd->rate & 3; - return (fd_wait_for_completion(jiffies + 2UL * HZ / 100, - (timeout_fn) floppy_ready)); + return fd_wait_for_completion(jiffies + 2UL * HZ / 100, + (timeout_fn)floppy_ready); } /* fdc_dtr */ static void tell_sector(void) { - printk(": track %d, head %d, sector %d, size %d", - R_TRACK, R_HEAD, R_SECTOR, R_SIZECODE); + pr_cont(": track %d, head %d, sector %d, size %d", + R_TRACK, R_HEAD, R_SECTOR, R_SIZECODE); } /* tell_sector */ +static void print_errors(void) +{ + DPRINT(""); + if (ST0 & ST0_ECE) { + pr_cont("Recalibrate failed!"); + } else if (ST2 & ST2_CRC) { + pr_cont("data CRC error"); + tell_sector(); + } else if (ST1 & ST1_CRC) { + pr_cont("CRC error"); + tell_sector(); + } else if ((ST1 & (ST1_MAM | ST1_ND)) || + (ST2 & ST2_MAM)) { + if (!probing) { + pr_cont("sector not found"); + tell_sector(); + } else + pr_cont("probe failed..."); + } else if (ST2 & ST2_WC) { /* seek error */ + pr_cont("wrong cylinder"); + } else if (ST2 & ST2_BC) { /* cylinder marked as bad */ + pr_cont("bad cylinder"); + } else { + pr_cont("unknown error. ST[0..2] are: 0x%x 0x%x 0x%x", + ST0, ST1, ST2); + tell_sector(); + } + pr_cont("\n"); +} + /* * OK, this error interpreting routine is called after a * DMA read/write has succeeded @@ -1437,7 +1445,7 @@ static int interpret_errors(void) char bad; if (inr != 7) { - DPRINT("-- FDC reply error"); + DPRINT("-- FDC reply error\n"); FDCS->reset = 1; return 1; } @@ -1450,43 +1458,17 @@ static int interpret_errors(void) bad = 1; if (ST1 & ST1_WP) { DPRINT("Drive is write protected\n"); - CLEARF(FD_DISK_WRITABLE); + clear_bit(FD_DISK_WRITABLE_BIT, &DRS->flags); cont->done(0); bad = 2; } else if (ST1 & ST1_ND) { - SETF(FD_NEED_TWADDLE); + set_bit(FD_NEED_TWADDLE_BIT, &DRS->flags); } else if (ST1 & ST1_OR) { if (DP->flags & FTD_MSG) DPRINT("Over/Underrun - retrying\n"); bad = 0; } else if (*errors >= DP->max_errors.reporting) { - DPRINT(""); - if (ST0 & ST0_ECE) { - printk("Recalibrate failed!"); - } else if (ST2 & ST2_CRC) { - printk("data CRC error"); - tell_sector(); - } else if (ST1 & ST1_CRC) { - printk("CRC error"); - tell_sector(); - } else if ((ST1 & (ST1_MAM | ST1_ND)) - || (ST2 & ST2_MAM)) { - if (!probing) { - printk("sector not found"); - tell_sector(); - } else - printk("probe failed..."); - } else if (ST2 & ST2_WC) { /* seek error */ - printk("wrong cylinder"); - } else if (ST2 & ST2_BC) { /* cylinder marked as bad */ - printk("bad cylinder"); - } else { - printk - ("unknown error. ST[0..2] are: 0x%x 0x%x 0x%x", - ST0, ST1, ST2); - tell_sector(); - } - printk("\n"); + print_errors(); } if (ST2 & ST2_WC || ST2 & ST2_BC) /* wrong cylinder => recal */ @@ -1531,9 +1513,9 @@ static void setup_rw_floppy(void) */ if (time_after(ready_date, jiffies + DP->select_delay)) { ready_date -= DP->select_delay; - function = (timeout_fn) floppy_start; + function = (timeout_fn)floppy_start; } else - function = (timeout_fn) setup_rw_floppy; + function = (timeout_fn)setup_rw_floppy; /* wait until the floppy is spinning fast enough */ if (fd_wait_for_completion(ready_date, function)) @@ -1551,7 +1533,7 @@ static void setup_rw_floppy(void) for (i = 0; i < raw_cmd->cmd_count; i++) r |= output_byte(raw_cmd->cmd[i]); - debugt("rw_command: "); + debugt(__func__, "rw_command"); if (r) { cont->error(); @@ -1574,7 +1556,7 @@ static int blind_seek; */ static void seek_interrupt(void) { - debugt("seek interrupt:"); + debugt(__func__, ""); if (inr != 2 || (ST0 & 0xF8) != 0x20) { DPRINT("seek failed\n"); DRS->track = NEED_2_RECAL; @@ -1583,14 +1565,11 @@ static void seek_interrupt(void) return; } if (DRS->track >= 0 && DRS->track != ST1 && !blind_seek) { -#ifdef DCL_DEBUG - if (DP->flags & FD_DEBUG) { - DPRINT - ("clearing NEWCHANGE flag because of effective seek\n"); - DPRINT("jiffies=%lu\n", jiffies); - } -#endif - CLEARF(FD_DISK_NEWCHANGE); /* effective seek */ + debug_dcl(DP->flags, + "clearing NEWCHANGE flag because of effective seek\n"); + debug_dcl(DP->flags, "jiffies=%lu\n", jiffies); + clear_bit(FD_DISK_NEWCHANGE_BIT, &DRS->flags); + /* effective seek */ DRS->select_date = jiffies; } DRS->track = ST1; @@ -1599,26 +1578,23 @@ static void seek_interrupt(void) static void check_wp(void) { - if (TESTF(FD_VERIFY)) { - /* check write protection */ + if (test_bit(FD_VERIFY_BIT, &DRS->flags)) { + /* check write protection */ output_byte(FD_GETSTATUS); output_byte(UNIT(current_drive)); if (result() != 1) { FDCS->reset = 1; return; } - CLEARF(FD_VERIFY); - CLEARF(FD_NEED_TWADDLE); -#ifdef DCL_DEBUG - if (DP->flags & FD_DEBUG) { - DPRINT("checking whether disk is write protected\n"); - DPRINT("wp=%x\n", ST3 & 0x40); - } -#endif + clear_bit(FD_VERIFY_BIT, &DRS->flags); + clear_bit(FD_NEED_TWADDLE_BIT, &DRS->flags); + debug_dcl(DP->flags, + "checking whether disk is write protected\n"); + debug_dcl(DP->flags, "wp=%x\n", ST3 & 0x40); if (!(ST3 & 0x40)) - SETF(FD_DISK_WRITABLE); + set_bit(FD_DISK_WRITABLE_BIT, &DRS->flags); else - CLEARF(FD_DISK_WRITABLE); + clear_bit(FD_DISK_WRITABLE_BIT, &DRS->flags); } } @@ -1628,19 +1604,15 @@ static void seek_floppy(void) blind_seek = 0; -#ifdef DCL_DEBUG - if (DP->flags & FD_DEBUG) { - DPRINT("calling disk change from seek\n"); - } -#endif + debug_dcl(DP->flags, "calling disk change from %s\n", __func__); - if (!TESTF(FD_DISK_NEWCHANGE) && + if (!test_bit(FD_DISK_NEWCHANGE_BIT, &DRS->flags) && disk_change(current_drive) && (raw_cmd->flags & FD_RAW_NEED_DISK)) { /* the media changed flag should be cleared after the seek. * If it isn't, this means that there is really no disk in * the drive. */ - SETF(FD_DISK_CHANGED); + set_bit(FD_DISK_CHANGED_BIT, &DRS->flags); cont->done(0); cont->redo(); return; @@ -1648,7 +1620,7 @@ static void seek_floppy(void) if (DRS->track <= NEED_1_RECAL) { recalibrate_floppy(); return; - } else if (TESTF(FD_DISK_NEWCHANGE) && + } else if (test_bit(FD_DISK_NEWCHANGE_BIT, &DRS->flags) && (raw_cmd->flags & FD_RAW_NEED_DISK) && (DRS->track <= NO_TRACK || DRS->track == raw_cmd->track)) { /* we seek to clear the media-changed condition. Does anybody @@ -1677,19 +1649,22 @@ static void seek_floppy(void) do_floppy = seek_interrupt; output_byte(FD_SEEK); output_byte(UNIT(current_drive)); - LAST_OUT(track); - debugt("seek command:"); + if (output_byte(track) < 0) { + reset_fdc(); + return; + } + debugt(__func__, ""); } static void recal_interrupt(void) { - debugt("recal interrupt:"); + debugt(__func__, ""); if (inr != 2) FDCS->reset = 1; else if (ST0 & ST0_ECE) { switch (DRS->track) { case NEED_1_RECAL: - debugt("recal interrupt need 1 recal:"); + debugt(__func__, "need 1 recal"); /* after a second recalibrate, we still haven't * reached track 0. Probably no drive. Raise an * error, as failing immediately might upset @@ -1698,25 +1673,21 @@ static void recal_interrupt(void) cont->redo(); return; case NEED_2_RECAL: - debugt("recal interrupt need 2 recal:"); + debugt(__func__, "need 2 recal"); /* If we already did a recalibrate, * and we are not at track 0, this * means we have moved. (The only way * not to move at recalibration is to * be already at track 0.) Clear the * new change flag */ -#ifdef DCL_DEBUG - if (DP->flags & FD_DEBUG) { - DPRINT - ("clearing NEWCHANGE flag because of second recalibrate\n"); - } -#endif + debug_dcl(DP->flags, + "clearing NEWCHANGE flag because of second recalibrate\n"); - CLEARF(FD_DISK_NEWCHANGE); + clear_bit(FD_DISK_NEWCHANGE_BIT, &DRS->flags); DRS->select_date = jiffies; /* fall through */ default: - debugt("recal interrupt default:"); + debugt(__func__, "default"); /* Recalibrate moves the head by at * most 80 steps. If after one * recalibrate we don't have reached @@ -1738,8 +1709,8 @@ static void print_result(char *message, int inr) DPRINT("%s ", message); if (inr >= 0) for (i = 0; i < inr; i++) - printk("repl[%d]=%x ", i, reply_buffer[i]); - printk("\n"); + pr_cont("repl[%d]=%x ", i, reply_buffer[i]); + pr_cont("\n"); } /* interrupt handler. Note that this can be called externally on the Sparc */ @@ -1760,10 +1731,10 @@ irqreturn_t floppy_interrupt(int irq, void *dev_id) do_floppy = NULL; if (fdc >= N_FDC || FDCS->address == -1) { /* we don't even know which FDC is the culprit */ - printk("DOR0=%x\n", fdc_state[0].dor); - printk("floppy interrupt on bizarre fdc %d\n", fdc); - printk("handler=%p\n", handler); - is_alive("bizarre fdc"); + pr_info("DOR0=%x\n", fdc_state[0].dor); + pr_info("floppy interrupt on bizarre fdc %d\n", fdc); + pr_info("handler=%pf\n", handler); + is_alive(__func__, "bizarre fdc"); return IRQ_NONE; } @@ -1777,7 +1748,7 @@ irqreturn_t floppy_interrupt(int irq, void *dev_id) * activity. */ - do_print = !handler && print_unex && !initialising; + do_print = !handler && print_unex && initialized; inr = result(); if (do_print) @@ -1790,15 +1761,15 @@ irqreturn_t floppy_interrupt(int irq, void *dev_id) if (do_print) print_result("sensei", inr); max_sensei--; - } while ((ST0 & 0x83) != UNIT(current_drive) && inr == 2 - && max_sensei); + } while ((ST0 & 0x83) != UNIT(current_drive) && + inr == 2 && max_sensei); } if (!handler) { FDCS->reset = 1; return IRQ_NONE; } schedule_bh(handler); - is_alive("normal interrupt end"); + is_alive(__func__, "normal interrupt end"); /* FIXME! Was it really for us? */ return IRQ_HANDLED; @@ -1806,10 +1777,11 @@ irqreturn_t floppy_interrupt(int irq, void *dev_id) static void recalibrate_floppy(void) { - debugt("recalibrate floppy:"); + debugt(__func__, ""); do_floppy = recal_interrupt; output_byte(FD_RECALIBRATE); - LAST_OUT(UNIT(current_drive)); + if (output_byte(UNIT(current_drive)) < 0) + reset_fdc(); } /* @@ -1817,10 +1789,10 @@ static void recalibrate_floppy(void) */ static void reset_interrupt(void) { - debugt("reset interrupt:"); + debugt(__func__, ""); result(); /* get the status ready for set_fdc */ if (FDCS->reset) { - printk("reset set in interrupt, calling %p\n", cont->error); + pr_info("reset set in interrupt, calling %pf\n", cont->error); cont->error(); /* a reset just after a reset. BAD! */ } cont->redo(); @@ -1858,53 +1830,49 @@ static void show_floppy(void) { int i; - printk("\n"); - printk("floppy driver state\n"); - printk("-------------------\n"); - printk("now=%lu last interrupt=%lu diff=%lu last called handler=%p\n", - jiffies, interruptjiffies, jiffies - interruptjiffies, - lasthandler); + pr_info("\n"); + pr_info("floppy driver state\n"); + pr_info("-------------------\n"); + pr_info("now=%lu last interrupt=%lu diff=%lu last called handler=%pf\n", + jiffies, interruptjiffies, jiffies - interruptjiffies, + lasthandler); -#ifdef FLOPPY_SANITY_CHECK - printk("timeout_message=%s\n", timeout_message); - printk("last output bytes:\n"); + pr_info("timeout_message=%s\n", timeout_message); + pr_info("last output bytes:\n"); for (i = 0; i < OLOGSIZE; i++) - printk("%2x %2x %lu\n", - output_log[(i + output_log_pos) % OLOGSIZE].data, - output_log[(i + output_log_pos) % OLOGSIZE].status, - output_log[(i + output_log_pos) % OLOGSIZE].jiffies); - printk("last result at %lu\n", resultjiffies); - printk("last redo_fd_request at %lu\n", lastredo); - for (i = 0; i < resultsize; i++) { - printk("%2x ", reply_buffer[i]); - } - printk("\n"); -#endif - - printk("status=%x\n", fd_inb(FD_STATUS)); - printk("fdc_busy=%lu\n", fdc_busy); + pr_info("%2x %2x %lu\n", + output_log[(i + output_log_pos) % OLOGSIZE].data, + output_log[(i + output_log_pos) % OLOGSIZE].status, + output_log[(i + output_log_pos) % OLOGSIZE].jiffies); + pr_info("last result at %lu\n", resultjiffies); + pr_info("last redo_fd_request at %lu\n", lastredo); + print_hex_dump(KERN_INFO, "", DUMP_PREFIX_NONE, 16, 1, + reply_buffer, resultsize, true); + + pr_info("status=%x\n", fd_inb(FD_STATUS)); + pr_info("fdc_busy=%lu\n", fdc_busy); if (do_floppy) - printk("do_floppy=%p\n", do_floppy); + pr_info("do_floppy=%pf\n", do_floppy); if (work_pending(&floppy_work)) - printk("floppy_work.func=%p\n", floppy_work.func); + pr_info("floppy_work.func=%pf\n", floppy_work.func); if (timer_pending(&fd_timer)) - printk("fd_timer.function=%p\n", fd_timer.function); + pr_info("fd_timer.function=%pf\n", fd_timer.function); if (timer_pending(&fd_timeout)) { - printk("timer_function=%p\n", fd_timeout.function); - printk("expires=%lu\n", fd_timeout.expires - jiffies); - printk("now=%lu\n", jiffies); - } - printk("cont=%p\n", cont); - printk("current_req=%p\n", current_req); - printk("command_status=%d\n", command_status); - printk("\n"); + pr_info("timer_function=%pf\n", fd_timeout.function); + pr_info("expires=%lu\n", fd_timeout.expires - jiffies); + pr_info("now=%lu\n", jiffies); + } + pr_info("cont=%p\n", cont); + pr_info("current_req=%p\n", current_req); + pr_info("command_status=%d\n", command_status); + pr_info("\n"); } static void floppy_shutdown(unsigned long data) { unsigned long flags; - if (!initialising) + if (initialized) show_floppy(); cancel_activity(); @@ -1916,17 +1884,17 @@ static void floppy_shutdown(unsigned long data) /* avoid dma going to a random drive after shutdown */ - if (!initialising) + if (initialized) DPRINT("floppy timeout called\n"); FDCS->reset = 1; if (cont) { cont->done(0); cont->redo(); /* this will recall reset when needed */ } else { - printk("no cont in shutdown!\n"); + pr_info("no cont in shutdown!\n"); process_fd_request(); } - is_alive("floppy shutdown"); + is_alive(__func__, ""); } /* start motor, check media-changed condition and write protection */ @@ -1954,27 +1922,26 @@ static int start_motor(void (*function)(void)) set_dor(fdc, mask, data); /* wait_for_completion also schedules reset if needed. */ - return (fd_wait_for_completion(DRS->select_date + DP->select_delay, - (timeout_fn) function)); + return fd_wait_for_completion(DRS->select_date + DP->select_delay, + (timeout_fn)function); } static void floppy_ready(void) { - CHECK_RESET; + if (FDCS->reset) { + reset_fdc(); + return; + } if (start_motor(floppy_ready)) return; if (fdc_dtr()) return; -#ifdef DCL_DEBUG - if (DP->flags & FD_DEBUG) { - DPRINT("calling disk change from floppy_ready\n"); - } -#endif + debug_dcl(DP->flags, "calling disk change from floppy_ready\n"); if (!(raw_cmd->flags & FD_RAW_NO_MOTOR) && disk_change(current_drive) && !DP->select_delay) - twaddle(); /* this clears the dcl on certain drive/controller - * combinations */ + twaddle(); /* this clears the dcl on certain + * drive/controller combinations */ #ifdef fd_chose_dma_mode if ((raw_cmd->flags & FD_RAW_READ) || (raw_cmd->flags & FD_RAW_WRITE)) { @@ -1998,15 +1965,11 @@ static void floppy_ready(void) static void floppy_start(void) { - reschedule_timeout(current_reqD, "floppy start", 0); + reschedule_timeout(current_reqD, "floppy start"); scandrives(); -#ifdef DCL_DEBUG - if (DP->flags & FD_DEBUG) { - DPRINT("setting NEWCHANGE in floppy_start\n"); - } -#endif - SETF(FD_DISK_NEWCHANGE); + debug_dcl(DP->flags, "setting NEWCHANGE in floppy_start\n"); + set_bit(FD_DISK_NEWCHANGE_BIT, &DRS->flags); floppy_ready(); } @@ -2026,7 +1989,7 @@ static void floppy_start(void) static void do_wakeup(void) { - reschedule_timeout(MAXTIMEOUT, "do wakeup", 0); + reschedule_timeout(MAXTIMEOUT, "do wakeup"); cont = NULL; command_status += 2; wake_up(&command_done); @@ -2046,7 +2009,7 @@ static struct cont_t intr_cont = { .done = (done_f)empty }; -static int wait_til_done(void (*handler)(void), int interruptible) +static int wait_til_done(void (*handler)(void), bool interruptible) { int ret; @@ -2064,7 +2027,7 @@ static int wait_til_done(void (*handler)(void), int interruptible) if (command_status >= 2 || !NO_SIGNAL) break; - is_alive("wait_til_done"); + is_alive(__func__, ""); schedule(); } @@ -2180,9 +2143,9 @@ static void format_interrupt(void) cont->redo(); } -#define CODE2SIZE (ssize = ((1 << SIZECODE) + 3) >> 2) -#define FM_MODE(x,y) ((y) & ~(((x)->rate & 0x80) >>1)) +#define FM_MODE(x, y) ((y) & ~(((x)->rate & 0x80) >> 1)) #define CT(x) ((x) | 0xc0) + static void setup_format_params(int track) { int n; @@ -2197,8 +2160,8 @@ static void setup_format_params(int track) raw_cmd = &default_raw_cmd; raw_cmd->track = track; - raw_cmd->flags = FD_RAW_WRITE | FD_RAW_INTR | FD_RAW_SPIN | - FD_RAW_NEED_DISK | FD_RAW_NEED_SEEK; + raw_cmd->flags = (FD_RAW_WRITE | FD_RAW_INTR | FD_RAW_SPIN | + FD_RAW_NEED_DISK | FD_RAW_NEED_SEEK); raw_cmd->rate = _floppy->rate & 0x43; raw_cmd->cmd_count = NR_F; COMMAND = FM_MODE(_floppy, FD_FORMAT); @@ -2257,7 +2220,7 @@ static void redo_format(void) buffer_track = -1; setup_format_params(format_req.track << STRETCH(_floppy)); floppy_start(); - debugt("queue format request"); + debugt(__func__, "queue format request"); } static struct cont_t format_cont = { @@ -2271,7 +2234,9 @@ static int do_format(int drive, struct format_descr *tmp_format_req) { int ret; - LOCK_FDC(drive, 1); + if (lock_fdc(drive, true)) + return -EINTR; + set_floppy(drive); if (!_floppy || _floppy->track > DP->tracks || @@ -2286,7 +2251,9 @@ static int do_format(int drive, struct format_descr *tmp_format_req) format_errors = 0; cont = &format_cont; errors = &format_errors; - IWAIT(redo_format); + ret = wait_til_done(redo_format, true); + if (ret == -EINTR) + return -EINTR; process_fd_request(); return ret; } @@ -2320,12 +2287,14 @@ static void request_done(int uptodate) struct request *req = current_req; unsigned long flags; int block; + char msg[sizeof("request done ") + sizeof(int) * 3]; probing = 0; - reschedule_timeout(MAXTIMEOUT, "request done %d", uptodate); + snprintf(msg, sizeof(msg), "request done %d", uptodate); + reschedule_timeout(MAXTIMEOUT, msg); if (!req) { - printk("floppy.c: no request in request_done\n"); + pr_info("floppy.c: no request in request_done\n"); return; } @@ -2377,7 +2346,7 @@ static void rw_interrupt(void) DRS->first_read_date = jiffies; nr_sectors = 0; - CODE2SIZE; + ssize = DIV_ROUND_UP(1 << SIZECODE, 4); if (ST1 & ST1_EOC) eoc = 1; @@ -2393,20 +2362,18 @@ static void rw_interrupt(void) R_HEAD - HEAD) * SECT_PER_TRACK + R_SECTOR - SECTOR + eoc) << SIZECODE >> 2; -#ifdef FLOPPY_SANITY_CHECK if (nr_sectors / ssize > DIV_ROUND_UP(in_sector_offset + current_count_sectors, ssize)) { DPRINT("long rw: %x instead of %lx\n", nr_sectors, current_count_sectors); - printk("rs=%d s=%d\n", R_SECTOR, SECTOR); - printk("rh=%d h=%d\n", R_HEAD, HEAD); - printk("rt=%d t=%d\n", R_TRACK, TRACK); - printk("heads=%d eoc=%d\n", heads, eoc); - printk("spt=%d st=%d ss=%d\n", SECT_PER_TRACK, - fsector_t, ssize); - printk("in_sector_offset=%d\n", in_sector_offset); + pr_info("rs=%d s=%d\n", R_SECTOR, SECTOR); + pr_info("rh=%d h=%d\n", R_HEAD, HEAD); + pr_info("rt=%d t=%d\n", R_TRACK, TRACK); + pr_info("heads=%d eoc=%d\n", heads, eoc); + pr_info("spt=%d st=%d ss=%d\n", + SECT_PER_TRACK, fsector_t, ssize); + pr_info("in_sector_offset=%d\n", in_sector_offset); } -#endif nr_sectors -= in_sector_offset; INFBOUND(nr_sectors, 0); @@ -2511,19 +2478,17 @@ static void copy_buffer(int ssize, int max_sector, int max_sector_2) blk_rq_sectors(current_req)); remaining = current_count_sectors << 9; -#ifdef FLOPPY_SANITY_CHECK if (remaining > blk_rq_bytes(current_req) && CT(COMMAND) == FD_WRITE) { DPRINT("in copy buffer\n"); - printk("current_count_sectors=%ld\n", current_count_sectors); - printk("remaining=%d\n", remaining >> 9); - printk("current_req->nr_sectors=%u\n", - blk_rq_sectors(current_req)); - printk("current_req->current_nr_sectors=%u\n", - blk_rq_cur_sectors(current_req)); - printk("max_sector=%d\n", max_sector); - printk("ssize=%d\n", ssize); + pr_info("current_count_sectors=%ld\n", current_count_sectors); + pr_info("remaining=%d\n", remaining >> 9); + pr_info("current_req->nr_sectors=%u\n", + blk_rq_sectors(current_req)); + pr_info("current_req->current_nr_sectors=%u\n", + blk_rq_cur_sectors(current_req)); + pr_info("max_sector=%d\n", max_sector); + pr_info("ssize=%d\n", ssize); } -#endif buffer_max = max(max_sector, buffer_max); @@ -2539,26 +2504,24 @@ static void copy_buffer(int ssize, int max_sector, int max_sector_2) SUPBOUND(size, remaining); buffer = page_address(bv->bv_page) + bv->bv_offset; -#ifdef FLOPPY_SANITY_CHECK if (dma_buffer + size > floppy_track_buffer + (max_buffer_sectors << 10) || dma_buffer < floppy_track_buffer) { DPRINT("buffer overrun in copy buffer %d\n", - (int)((floppy_track_buffer - - dma_buffer) >> 9)); - printk("fsector_t=%d buffer_min=%d\n", - fsector_t, buffer_min); - printk("current_count_sectors=%ld\n", - current_count_sectors); + (int)((floppy_track_buffer - dma_buffer) >> 9)); + pr_info("fsector_t=%d buffer_min=%d\n", + fsector_t, buffer_min); + pr_info("current_count_sectors=%ld\n", + current_count_sectors); if (CT(COMMAND) == FD_READ) - printk("read\n"); + pr_info("read\n"); if (CT(COMMAND) == FD_WRITE) - printk("write\n"); + pr_info("write\n"); break; } if (((unsigned long)buffer) % 512) DPRINT("%p buffer not aligned\n", buffer); -#endif + if (CT(COMMAND) == FD_READ) memcpy(buffer, dma_buffer, size); else @@ -2567,13 +2530,11 @@ static void copy_buffer(int ssize, int max_sector, int max_sector_2) remaining -= size; dma_buffer += size; } -#ifdef FLOPPY_SANITY_CHECK if (remaining) { if (remaining > 0) max_sector -= remaining >> 9; DPRINT("weirdness: remaining %d\n", remaining >> 9); } -#endif } /* work around a bug in pseudo DMA @@ -2593,15 +2554,14 @@ static void virtualdmabug_workaround(void) hard_sectors = raw_cmd->length >> (7 + SIZECODE); end_sector = SECTOR + hard_sectors - 1; -#ifdef FLOPPY_SANITY_CHECK if (end_sector > SECT_PER_TRACK) { - printk("too many sectors %d > %d\n", - end_sector, SECT_PER_TRACK); + pr_info("too many sectors %d > %d\n", + end_sector, SECT_PER_TRACK); return; } -#endif - SECT_PER_TRACK = end_sector; /* make sure SECT_PER_TRACK points - * to end of transfer */ + SECT_PER_TRACK = end_sector; + /* make sure SECT_PER_TRACK + * points to end of transfer */ } } @@ -2624,7 +2584,7 @@ static int make_raw_rw_request(void) int ssize; if (max_buffer_sectors == 0) { - printk("VFS: Block I/O scheduled on unopened device\n"); + pr_info("VFS: Block I/O scheduled on unopened device\n"); return 0; } @@ -2641,7 +2601,7 @@ static int make_raw_rw_request(void) raw_cmd->flags |= FD_RAW_WRITE; COMMAND = FM_MODE(_floppy, FD_WRITE); } else { - DPRINT("make_raw_rw_request: unknown command\n"); + DPRINT("%s: unknown command\n", __func__); return 0; } @@ -2659,7 +2619,8 @@ static int make_raw_rw_request(void) HEAD = fsector_t / _floppy->sect; if (((_floppy->stretch & (FD_SWAPSIDES | FD_SECTBASEMASK)) || - TESTF(FD_NEED_TWADDLE)) && fsector_t < _floppy->sect) + test_bit(FD_NEED_TWADDLE_BIT, &DRS->flags)) && + fsector_t < _floppy->sect) max_sector = _floppy->sect; /* 2M disks have phantom sectors on the first track */ @@ -2685,7 +2646,7 @@ static int make_raw_rw_request(void) raw_cmd->track = TRACK << STRETCH(_floppy); DR_SELECT = UNIT(current_drive) + PH_HEAD(_floppy, HEAD); GAP = _floppy->gap; - CODE2SIZE; + ssize = DIV_ROUND_UP(1 << SIZECODE, 4); SECT_PER_TRACK = _floppy->sect << 2 >> SIZECODE; SECTOR = ((fsector_t % _floppy->sect) << 2 >> SIZECODE) + FD_SECTBASE(_floppy); @@ -2730,8 +2691,10 @@ static int make_raw_rw_request(void) } } else if (in_sector_offset || blk_rq_sectors(current_req) < ssize) { if (CT(COMMAND) == FD_WRITE) { - if (fsector_t + blk_rq_sectors(current_req) > ssize && - fsector_t + blk_rq_sectors(current_req) < ssize + ssize) + unsigned int sectors; + + sectors = fsector_t + blk_rq_sectors(current_req); + if (sectors > ssize && sectors < ssize + ssize) max_size = ssize + ssize; else max_size = ssize; @@ -2752,12 +2715,10 @@ static int make_raw_rw_request(void) * on a 64 bit machine! */ max_size = buffer_chain_size(); - dma_limit = - (MAX_DMA_ADDRESS - - ((unsigned long)current_req->buffer)) >> 9; - if ((unsigned long)max_size > dma_limit) { + dma_limit = (MAX_DMA_ADDRESS - + ((unsigned long)current_req->buffer)) >> 9; + if ((unsigned long)max_size > dma_limit) max_size = dma_limit; - } /* 64 kb boundaries */ if (CROSS_64KB(current_req->buffer, max_size << 9)) max_size = (K_64 - @@ -2773,16 +2734,16 @@ static int make_raw_rw_request(void) */ if (!direct || (indirect * 2 > direct * 3 && - *errors < DP->max_errors.read_track && ((!probing - || (DP->read_track & (1 << DRS->probed_format)))))) { + *errors < DP->max_errors.read_track && + ((!probing || + (DP->read_track & (1 << DRS->probed_format)))))) { max_size = blk_rq_sectors(current_req); } else { raw_cmd->kernel_data = current_req->buffer; raw_cmd->length = current_count_sectors << 9; if (raw_cmd->length == 0) { - DPRINT - ("zero dma transfer attempted from make_raw_request\n"); - DPRINT("indirect=%d direct=%d fsector_t=%d", + DPRINT("%s: zero dma transfer attempted\n", __func__); + DPRINT("indirect=%d direct=%d fsector_t=%d\n", indirect, direct, fsector_t); return 0; } @@ -2802,25 +2763,22 @@ static int make_raw_rw_request(void) ((CT(COMMAND) == FD_READ || (!in_sector_offset && blk_rq_sectors(current_req) >= ssize)) && max_sector > 2 * max_buffer_sectors + buffer_min && - max_size + fsector_t > 2 * max_buffer_sectors + buffer_min) - /* not enough space */ - ) { + max_size + fsector_t > 2 * max_buffer_sectors + buffer_min)) { + /* not enough space */ buffer_track = -1; buffer_drive = current_drive; buffer_max = buffer_min = aligned_sector_t; } raw_cmd->kernel_data = floppy_track_buffer + - ((aligned_sector_t - buffer_min) << 9); + ((aligned_sector_t - buffer_min) << 9); if (CT(COMMAND) == FD_WRITE) { /* copy write buffer to track buffer. * if we get here, we know that the write * is either aligned or the data already in the buffer * (buffer will be overwritten) */ -#ifdef FLOPPY_SANITY_CHECK if (in_sector_offset && buffer_track == -1) DPRINT("internal error offset !=0 on write\n"); -#endif buffer_track = raw_cmd->track; buffer_drive = current_drive; copy_buffer(ssize, max_sector, @@ -2834,7 +2792,6 @@ static int make_raw_rw_request(void) raw_cmd->length = in_sector_offset + current_count_sectors; raw_cmd->length = ((raw_cmd->length - 1) | (ssize - 1)) + 1; raw_cmd->length <<= 9; -#ifdef FLOPPY_SANITY_CHECK if ((raw_cmd->length < current_count_sectors << 9) || (raw_cmd->kernel_data != current_req->buffer && CT(COMMAND) == FD_WRITE && @@ -2845,19 +2802,19 @@ static int make_raw_rw_request(void) DPRINT("fractionary current count b=%lx s=%lx\n", raw_cmd->length, current_count_sectors); if (raw_cmd->kernel_data != current_req->buffer) - printk("addr=%d, length=%ld\n", - (int)((raw_cmd->kernel_data - - floppy_track_buffer) >> 9), - current_count_sectors); - printk("st=%d ast=%d mse=%d msi=%d\n", - fsector_t, aligned_sector_t, max_sector, max_size); - printk("ssize=%x SIZECODE=%d\n", ssize, SIZECODE); - printk("command=%x SECTOR=%d HEAD=%d, TRACK=%d\n", - COMMAND, SECTOR, HEAD, TRACK); - printk("buffer drive=%d\n", buffer_drive); - printk("buffer track=%d\n", buffer_track); - printk("buffer_min=%d\n", buffer_min); - printk("buffer_max=%d\n", buffer_max); + pr_info("addr=%d, length=%ld\n", + (int)((raw_cmd->kernel_data - + floppy_track_buffer) >> 9), + current_count_sectors); + pr_info("st=%d ast=%d mse=%d msi=%d\n", + fsector_t, aligned_sector_t, max_sector, max_size); + pr_info("ssize=%x SIZECODE=%d\n", ssize, SIZECODE); + pr_info("command=%x SECTOR=%d HEAD=%d, TRACK=%d\n", + COMMAND, SECTOR, HEAD, TRACK); + pr_info("buffer drive=%d\n", buffer_drive); + pr_info("buffer track=%d\n", buffer_track); + pr_info("buffer_min=%d\n", buffer_min); + pr_info("buffer_max=%d\n", buffer_max); return 0; } @@ -2868,14 +2825,14 @@ static int make_raw_rw_request(void) raw_cmd->kernel_data + raw_cmd->length > floppy_track_buffer + (max_buffer_sectors << 10)) { DPRINT("buffer overrun in schedule dma\n"); - printk("fsector_t=%d buffer_min=%d current_count=%ld\n", - fsector_t, buffer_min, raw_cmd->length >> 9); - printk("current_count_sectors=%ld\n", - current_count_sectors); + pr_info("fsector_t=%d buffer_min=%d current_count=%ld\n", + fsector_t, buffer_min, raw_cmd->length >> 9); + pr_info("current_count_sectors=%ld\n", + current_count_sectors); if (CT(COMMAND) == FD_READ) - printk("read\n"); + pr_info("read\n"); if (CT(COMMAND) == FD_WRITE) - printk("write\n"); + pr_info("write\n"); return 0; } } else if (raw_cmd->length > blk_rq_bytes(current_req) || @@ -2884,14 +2841,13 @@ static int make_raw_rw_request(void) return 0; } else if (raw_cmd->length < current_count_sectors << 9) { DPRINT("more sectors than bytes\n"); - printk("bytes=%ld\n", raw_cmd->length >> 9); - printk("sectors=%ld\n", current_count_sectors); + pr_info("bytes=%ld\n", raw_cmd->length >> 9); + pr_info("sectors=%ld\n", current_count_sectors); } if (raw_cmd->length == 0) { DPRINT("zero dma transfer attempted from make_raw_request\n"); return 0; } -#endif virtualdmabug_workaround(); return 2; @@ -2899,7 +2855,6 @@ static int make_raw_rw_request(void) static void redo_fd_request(void) { -#define REPEAT {request_done(0); continue; } int drive; int tmp; @@ -2907,63 +2862,63 @@ static void redo_fd_request(void) if (current_drive < N_DRIVE) floppy_off(current_drive); - for (;;) { - if (!current_req) { - struct request *req; - - spin_lock_irq(floppy_queue->queue_lock); - req = blk_fetch_request(floppy_queue); - spin_unlock_irq(floppy_queue->queue_lock); - if (!req) { - do_floppy = NULL; - unlock_fdc(); - return; - } - current_req = req; - } - drive = (long)current_req->rq_disk->private_data; - set_fdc(drive); - reschedule_timeout(current_reqD, "redo fd request", 0); +do_request: + if (!current_req) { + struct request *req; - set_floppy(drive); - raw_cmd = &default_raw_cmd; - raw_cmd->flags = 0; - if (start_motor(redo_fd_request)) + spin_lock_irq(floppy_queue->queue_lock); + req = blk_fetch_request(floppy_queue); + spin_unlock_irq(floppy_queue->queue_lock); + if (!req) { + do_floppy = NULL; + unlock_fdc(); return; - disk_change(current_drive); - if (test_bit(current_drive, &fake_change) || - TESTF(FD_DISK_CHANGED)) { - DPRINT("disk absent or changed during operation\n"); - REPEAT; - } - if (!_floppy) { /* Autodetection */ - if (!probing) { - DRS->probed_format = 0; - if (next_valid_format()) { - DPRINT("no autodetectable formats\n"); - _floppy = NULL; - REPEAT; - } - } - probing = 1; - _floppy = - floppy_type + DP->autodetect[DRS->probed_format]; - } else - probing = 0; - errors = &(current_req->errors); - tmp = make_raw_rw_request(); - if (tmp < 2) { - request_done(tmp); - continue; } + current_req = req; + } + drive = (long)current_req->rq_disk->private_data; + set_fdc(drive); + reschedule_timeout(current_reqD, "redo fd request"); - if (TESTF(FD_NEED_TWADDLE)) - twaddle(); - schedule_bh(floppy_start); - debugt("queue fd request"); + set_floppy(drive); + raw_cmd = &default_raw_cmd; + raw_cmd->flags = 0; + if (start_motor(redo_fd_request)) return; + + disk_change(current_drive); + if (test_bit(current_drive, &fake_change) || + test_bit(FD_DISK_CHANGED_BIT, &DRS->flags)) { + DPRINT("disk absent or changed during operation\n"); + request_done(0); + goto do_request; + } + if (!_floppy) { /* Autodetection */ + if (!probing) { + DRS->probed_format = 0; + if (next_valid_format()) { + DPRINT("no autodetectable formats\n"); + _floppy = NULL; + request_done(0); + goto do_request; + } + } + probing = 1; + _floppy = floppy_type + DP->autodetect[DRS->probed_format]; + } else + probing = 0; + errors = &(current_req->errors); + tmp = make_raw_rw_request(); + if (tmp < 2) { + request_done(tmp); + goto do_request; } -#undef REPEAT + + if (test_bit(FD_NEED_TWADDLE_BIT, &DRS->flags)) + twaddle(); + schedule_bh(floppy_start); + debugt(__func__, "queue fd request"); + return; } static struct cont_t rw_cont = { @@ -2979,30 +2934,30 @@ static void process_fd_request(void) schedule_bh(redo_fd_request); } -static void do_fd_request(struct request_queue * q) +static void do_fd_request(struct request_queue *q) { if (max_buffer_sectors == 0) { - printk("VFS: do_fd_request called on non-open device\n"); + pr_info("VFS: %s called on non-open device\n", __func__); return; } if (usage_count == 0) { - printk("warning: usage count=0, current_req=%p exiting\n", - current_req); - printk("sect=%ld type=%x flags=%x\n", - (long)blk_rq_pos(current_req), current_req->cmd_type, - current_req->cmd_flags); + pr_info("warning: usage count=0, current_req=%p exiting\n", + current_req); + pr_info("sect=%ld type=%x flags=%x\n", + (long)blk_rq_pos(current_req), current_req->cmd_type, + current_req->cmd_flags); return; } if (test_bit(0, &fdc_busy)) { /* fdc busy, this new request will be treated when the current one is done */ - is_alive("do fd request, old request running"); + is_alive(__func__, "old request running"); return; } - lock_fdc(MAXTIMEOUT, 0); + lock_fdc(MAXTIMEOUT, false); process_fd_request(); - is_alive("do fd request"); + is_alive(__func__, ""); } static struct cont_t poll_cont = { @@ -3012,24 +2967,18 @@ static struct cont_t poll_cont = { .done = generic_done }; -static int poll_drive(int interruptible, int flag) +static int poll_drive(bool interruptible, int flag) { - int ret; - /* no auto-sense, just clear dcl */ raw_cmd = &default_raw_cmd; raw_cmd->flags = flag; raw_cmd->track = 0; raw_cmd->cmd_count = 0; cont = &poll_cont; -#ifdef DCL_DEBUG - if (DP->flags & FD_DEBUG) { - DPRINT("setting NEWCHANGE in poll_drive\n"); - } -#endif - SETF(FD_DISK_NEWCHANGE); - WAIT(floppy_ready); - return ret; + debug_dcl(DP->flags, "setting NEWCHANGE in poll_drive\n"); + set_bit(FD_DISK_NEWCHANGE_BIT, &DRS->flags); + + return wait_til_done(floppy_ready, interruptible); } /* @@ -3039,7 +2988,7 @@ static int poll_drive(int interruptible, int flag) static void reset_intr(void) { - printk("weird, reset interrupt called\n"); + pr_info("weird, reset interrupt called\n"); } static struct cont_t reset_cont = { @@ -3049,20 +2998,23 @@ static struct cont_t reset_cont = { .done = generic_done }; -static int user_reset_fdc(int drive, int arg, int interruptible) +static int user_reset_fdc(int drive, int arg, bool interruptible) { int ret; - ret = 0; - LOCK_FDC(drive, interruptible); + if (lock_fdc(drive, interruptible)) + return -EINTR; + if (arg == FD_RESET_ALWAYS) FDCS->reset = 1; if (FDCS->reset) { cont = &reset_cont; - WAIT(reset_fdc); + ret = wait_til_done(reset_fdc, interruptible); + if (ret == -EINTR) + return -EINTR; } process_fd_request(); - return ret; + return 0; } /* @@ -3075,17 +3027,12 @@ static inline int fd_copyout(void __user *param, const void *address, return copy_to_user(param, address, size) ? -EFAULT : 0; } -static inline int fd_copyin(void __user *param, void *address, unsigned long size) +static inline int fd_copyin(void __user *param, void *address, + unsigned long size) { return copy_from_user(address, param, size) ? -EFAULT : 0; } -#define _COPYOUT(x) (copy_to_user((void __user *)param, &(x), sizeof(x)) ? -EFAULT : 0) -#define _COPYIN(x) (copy_from_user(&(x), (void __user *)param, sizeof(x)) ? -EFAULT : 0) - -#define COPYOUT(x) ECALL(_COPYOUT(x)) -#define COPYIN(x) ECALL(_COPYIN(x)) - static inline const char *drive_name(int type, int drive) { struct floppy_struct *floppy; @@ -3156,23 +3103,29 @@ static struct cont_t raw_cmd_cont = { .done = raw_cmd_done }; -static inline int raw_cmd_copyout(int cmd, char __user *param, +static inline int raw_cmd_copyout(int cmd, void __user *param, struct floppy_raw_cmd *ptr) { int ret; while (ptr) { - COPYOUT(*ptr); + ret = copy_to_user(param, ptr, sizeof(*ptr)); + if (ret) + return -EFAULT; param += sizeof(struct floppy_raw_cmd); if ((ptr->flags & FD_RAW_READ) && ptr->buffer_length) { - if (ptr->length >= 0 - && ptr->length <= ptr->buffer_length) - ECALL(fd_copyout - (ptr->data, ptr->kernel_data, - ptr->buffer_length - ptr->length)); + if (ptr->length >= 0 && + ptr->length <= ptr->buffer_length) { + long length = ptr->buffer_length - ptr->length; + ret = fd_copyout(ptr->data, ptr->kernel_data, + length); + if (ret) + return ret; + } } ptr = ptr->next; } + return 0; } @@ -3195,7 +3148,7 @@ static void raw_cmd_free(struct floppy_raw_cmd **ptr) } } -static inline int raw_cmd_copyin(int cmd, char __user *param, +static inline int raw_cmd_copyin(int cmd, void __user *param, struct floppy_raw_cmd **rcmd) { struct floppy_raw_cmd *ptr; @@ -3203,17 +3156,19 @@ static inline int raw_cmd_copyin(int cmd, char __user *param, int i; *rcmd = NULL; - while (1) { - ptr = (struct floppy_raw_cmd *) - kmalloc(sizeof(struct floppy_raw_cmd), GFP_USER); - if (!ptr) - return -ENOMEM; - *rcmd = ptr; - COPYIN(*ptr); - ptr->next = NULL; - ptr->buffer_length = 0; - param += sizeof(struct floppy_raw_cmd); - if (ptr->cmd_count > 33) + +loop: + ptr = kmalloc(sizeof(struct floppy_raw_cmd), GFP_USER); + if (!ptr) + return -ENOMEM; + *rcmd = ptr; + ret = copy_from_user(ptr, param, sizeof(*ptr)); + if (ret) + return -EFAULT; + ptr->next = NULL; + ptr->buffer_length = 0; + param += sizeof(struct floppy_raw_cmd); + if (ptr->cmd_count > 33) /* the command may now also take up the space * initially intended for the reply & the * reply count. Needed for long 82078 commands @@ -3222,31 +3177,35 @@ static inline int raw_cmd_copyin(int cmd, char __user *param, * 16 bytes for a structure, you'll one day * discover that you really need 17... */ + return -EINVAL; + + for (i = 0; i < 16; i++) + ptr->reply[i] = 0; + ptr->resultcode = 0; + ptr->kernel_data = NULL; + + if (ptr->flags & (FD_RAW_READ | FD_RAW_WRITE)) { + if (ptr->length <= 0) return -EINVAL; + ptr->kernel_data = (char *)fd_dma_mem_alloc(ptr->length); + fallback_on_nodma_alloc(&ptr->kernel_data, ptr->length); + if (!ptr->kernel_data) + return -ENOMEM; + ptr->buffer_length = ptr->length; + } + if (ptr->flags & FD_RAW_WRITE) { + ret = fd_copyin(ptr->data, ptr->kernel_data, ptr->length); + if (ret) + return ret; + } - for (i = 0; i < 16; i++) - ptr->reply[i] = 0; - ptr->resultcode = 0; - ptr->kernel_data = NULL; - - if (ptr->flags & (FD_RAW_READ | FD_RAW_WRITE)) { - if (ptr->length <= 0) - return -EINVAL; - ptr->kernel_data = - (char *)fd_dma_mem_alloc(ptr->length); - fallback_on_nodma_alloc(&ptr->kernel_data, ptr->length); - if (!ptr->kernel_data) - return -ENOMEM; - ptr->buffer_length = ptr->length; - } - if (ptr->flags & FD_RAW_WRITE) - ECALL(fd_copyin(ptr->data, ptr->kernel_data, - ptr->length)); + if (ptr->flags & FD_RAW_MORE) { rcmd = &(ptr->next); - if (!(ptr->flags & FD_RAW_MORE)) - return 0; ptr->rate &= 0x43; + goto loop; } + + return 0; } static int raw_cmd_ioctl(int cmd, void __user *param) @@ -3283,12 +3242,8 @@ static int raw_cmd_ioctl(int cmd, void __user *param) raw_cmd = my_raw_cmd; cont = &raw_cmd_cont; - ret = wait_til_done(floppy_start, 1); -#ifdef DCL_DEBUG - if (DP->flags & FD_DEBUG) { - DPRINT("calling disk change from raw_cmd ioctl\n"); - } -#endif + ret = wait_til_done(floppy_start, true); + debug_dcl(DP->flags, "calling disk change from raw_cmd ioctl\n"); if (ret != -EINTR && FDCS->reset) ret = -EIO; @@ -3327,7 +3282,7 @@ static inline int set_geometry(unsigned int cmd, struct floppy_struct *g, if (!capable(CAP_SYS_ADMIN)) return -EPERM; mutex_lock(&open_lock); - if (lock_fdc(drive, 1)) { + if (lock_fdc(drive, true)) { mutex_unlock(&open_lock); return -EINTR; } @@ -3346,11 +3301,15 @@ static inline int set_geometry(unsigned int cmd, struct floppy_struct *g, mutex_unlock(&open_lock); } else { int oldStretch; - LOCK_FDC(drive, 1); - if (cmd != FDDEFPRM) + + if (lock_fdc(drive, true)) + return -EINTR; + if (cmd != FDDEFPRM) { /* notice a disk change immediately, else * we lose our settings immediately*/ - CALL(poll_drive(1, FD_RAW_NEED_DISK)); + if (poll_drive(true, FD_RAW_NEED_DISK) == -EINTR) + return -EINTR; + } oldStretch = g->stretch; user_params[drive] = *g; if (buffer_drive == drive) @@ -3415,7 +3374,7 @@ static inline int normalize_ioctl(int *cmd, int *size) *size = _IOC_SIZE(*cmd); *cmd = ioctl_table[i]; if (*size > _IOC_SIZE(*cmd)) { - printk("ioctl not yet supported\n"); + pr_info("ioctl not yet supported\n"); return -EFAULT; } return 0; @@ -3429,8 +3388,10 @@ static int get_floppy_geometry(int drive, int type, struct floppy_struct **g) if (type) *g = &floppy_type[type]; else { - LOCK_FDC(drive, 0); - CALL(poll_drive(0, 0)); + if (lock_fdc(drive, false)) + return -EINTR; + if (poll_drive(false, 0) == -EINTR) + return -EINTR; process_fd_request(); *g = current_type[drive]; } @@ -3459,10 +3420,6 @@ static int fd_getgeo(struct block_device *bdev, struct hd_geometry *geo) static int fd_ioctl(struct block_device *bdev, fmode_t mode, unsigned int cmd, unsigned long param) { -#define FD_IOCTL_ALLOWED (mode & (FMODE_WRITE|FMODE_WRITE_IOCTL)) -#define OUT(c,x) case c: outparam = (const char *) (x); break -#define IN(c,x,tag) case c: *(x) = inparam. tag ; return 0 - int drive = (long)bdev->bd_disk->private_data; int type = ITYPE(UDRS->fd_device); int i; @@ -3474,26 +3431,28 @@ static int fd_ioctl(struct block_device *bdev, fmode_t mode, unsigned int cmd, struct floppy_max_errors max_errors; struct floppy_drive_params dp; } inparam; /* parameters coming from user space */ - const char *outparam; /* parameters passed back to user space */ + const void *outparam; /* parameters passed back to user space */ /* convert compatibility eject ioctls into floppy eject ioctl. * We do this in order to provide a means to eject floppy disks before * installing the new fdutils package */ if (cmd == CDROMEJECT || /* CD-ROM eject */ - cmd == 0x6470 /* SunOS floppy eject */ ) { + cmd == 0x6470) { /* SunOS floppy eject */ DPRINT("obsolete eject ioctl\n"); DPRINT("please use floppycontrol --eject\n"); cmd = FDEJECT; } - /* convert the old style command into a new style command */ - if ((cmd & 0xff00) == 0x0200) { - ECALL(normalize_ioctl(&cmd, &size)); - } else + if (!((cmd & 0xff00) == 0x0200)) return -EINVAL; + /* convert the old style command into a new style command */ + ret = normalize_ioctl(&cmd, &size); + if (ret) + return ret; + /* permission checks */ - if (((cmd & 0x40) && !FD_IOCTL_ALLOWED) || + if (((cmd & 0x40) && !(mode & (FMODE_WRITE | FMODE_WRITE_IOCTL))) || ((cmd & 0x80) && !capable(CAP_SYS_ADMIN))) return -EPERM; @@ -3501,129 +3460,142 @@ static int fd_ioctl(struct block_device *bdev, fmode_t mode, unsigned int cmd, return -EINVAL; /* copyin */ - CLEARSTRUCT(&inparam); - if (_IOC_DIR(cmd) & _IOC_WRITE) - ECALL(fd_copyin((void __user *)param, &inparam, size)) - - switch (cmd) { - case FDEJECT: - if (UDRS->fd_ref != 1) - /* somebody else has this drive open */ - return -EBUSY; - LOCK_FDC(drive, 1); - - /* do the actual eject. Fails on - * non-Sparc architectures */ - ret = fd_eject(UNIT(drive)); - - USETF(FD_DISK_CHANGED); - USETF(FD_VERIFY); - process_fd_request(); + memset(&inparam, 0, sizeof(inparam)); + if (_IOC_DIR(cmd) & _IOC_WRITE) { + ret = fd_copyin((void __user *)param, &inparam, size); + if (ret) return ret; - case FDCLRPRM: - LOCK_FDC(drive, 1); - current_type[drive] = NULL; - floppy_sizes[drive] = MAX_DISK_SIZE << 1; - UDRS->keep_data = 0; - return invalidate_drive(bdev); - case FDSETPRM: - case FDDEFPRM: - return set_geometry(cmd, &inparam.g, - drive, type, bdev); - case FDGETPRM: - ECALL(get_floppy_geometry(drive, type, - (struct floppy_struct **) - &outparam)); - break; - - case FDMSGON: - UDP->flags |= FTD_MSG; - return 0; - case FDMSGOFF: - UDP->flags &= ~FTD_MSG; - return 0; - - case FDFMTBEG: - LOCK_FDC(drive, 1); - CALL(poll_drive(1, FD_RAW_NEED_DISK)); - ret = UDRS->flags; - process_fd_request(); - if (ret & FD_VERIFY) - return -ENODEV; - if (!(ret & FD_DISK_WRITABLE)) - return -EROFS; - return 0; - case FDFMTTRK: - if (UDRS->fd_ref != 1) - return -EBUSY; - return do_format(drive, &inparam.f); - case FDFMTEND: - case FDFLUSH: - LOCK_FDC(drive, 1); - return invalidate_drive(bdev); - - case FDSETEMSGTRESH: - UDP->max_errors.reporting = - (unsigned short)(param & 0x0f); - return 0; - OUT(FDGETMAXERRS, &UDP->max_errors); - IN(FDSETMAXERRS, &UDP->max_errors, max_errors); - - case FDGETDRVTYP: - outparam = drive_name(type, drive); - SUPBOUND(size, strlen(outparam) + 1); - break; - - IN(FDSETDRVPRM, UDP, dp); - OUT(FDGETDRVPRM, UDP); - - case FDPOLLDRVSTAT: - LOCK_FDC(drive, 1); - CALL(poll_drive(1, FD_RAW_NEED_DISK)); - process_fd_request(); - /* fall through */ - OUT(FDGETDRVSTAT, UDRS); - - case FDRESET: - return user_reset_fdc(drive, (int)param, 1); - - OUT(FDGETFDCSTAT, UFDCS); + } - case FDWERRORCLR: - CLEARSTRUCT(UDRWE); - return 0; - OUT(FDWERRORGET, UDRWE); - - case FDRAWCMD: - if (type) - return -EINVAL; - LOCK_FDC(drive, 1); - set_floppy(drive); - CALL(i = raw_cmd_ioctl(cmd, (void __user *)param)); - process_fd_request(); - return i; + switch (cmd) { + case FDEJECT: + if (UDRS->fd_ref != 1) + /* somebody else has this drive open */ + return -EBUSY; + if (lock_fdc(drive, true)) + return -EINTR; - case FDTWADDLE: - LOCK_FDC(drive, 1); - twaddle(); - process_fd_request(); - return 0; + /* do the actual eject. Fails on + * non-Sparc architectures */ + ret = fd_eject(UNIT(drive)); - default: + set_bit(FD_DISK_CHANGED_BIT, &UDRS->flags); + set_bit(FD_VERIFY_BIT, &UDRS->flags); + process_fd_request(); + return ret; + case FDCLRPRM: + if (lock_fdc(drive, true)) + return -EINTR; + current_type[drive] = NULL; + floppy_sizes[drive] = MAX_DISK_SIZE << 1; + UDRS->keep_data = 0; + return invalidate_drive(bdev); + case FDSETPRM: + case FDDEFPRM: + return set_geometry(cmd, &inparam.g, drive, type, bdev); + case FDGETPRM: + ret = get_floppy_geometry(drive, type, + (struct floppy_struct **)&outparam); + if (ret) + return ret; + break; + case FDMSGON: + UDP->flags |= FTD_MSG; + return 0; + case FDMSGOFF: + UDP->flags &= ~FTD_MSG; + return 0; + case FDFMTBEG: + if (lock_fdc(drive, true)) + return -EINTR; + if (poll_drive(true, FD_RAW_NEED_DISK) == -EINTR) + return -EINTR; + ret = UDRS->flags; + process_fd_request(); + if (ret & FD_VERIFY) + return -ENODEV; + if (!(ret & FD_DISK_WRITABLE)) + return -EROFS; + return 0; + case FDFMTTRK: + if (UDRS->fd_ref != 1) + return -EBUSY; + return do_format(drive, &inparam.f); + case FDFMTEND: + case FDFLUSH: + if (lock_fdc(drive, true)) + return -EINTR; + return invalidate_drive(bdev); + case FDSETEMSGTRESH: + UDP->max_errors.reporting = (unsigned short)(param & 0x0f); + return 0; + case FDGETMAXERRS: + outparam = &UDP->max_errors; + break; + case FDSETMAXERRS: + UDP->max_errors = inparam.max_errors; + break; + case FDGETDRVTYP: + outparam = drive_name(type, drive); + SUPBOUND(size, strlen((const char *)outparam) + 1); + break; + case FDSETDRVPRM: + *UDP = inparam.dp; + break; + case FDGETDRVPRM: + outparam = UDP; + break; + case FDPOLLDRVSTAT: + if (lock_fdc(drive, true)) + return -EINTR; + if (poll_drive(true, FD_RAW_NEED_DISK) == -EINTR) + return -EINTR; + process_fd_request(); + /* fall through */ + case FDGETDRVSTAT: + outparam = UDRS; + break; + case FDRESET: + return user_reset_fdc(drive, (int)param, true); + case FDGETFDCSTAT: + outparam = UFDCS; + break; + case FDWERRORCLR: + memset(UDRWE, 0, sizeof(*UDRWE)); + return 0; + case FDWERRORGET: + outparam = UDRWE; + break; + case FDRAWCMD: + if (type) return -EINVAL; - } + if (lock_fdc(drive, true)) + return -EINTR; + set_floppy(drive); + i = raw_cmd_ioctl(cmd, (void __user *)param); + if (i == -EINTR) + return -EINTR; + process_fd_request(); + return i; + case FDTWADDLE: + if (lock_fdc(drive, true)) + return -EINTR; + twaddle(); + process_fd_request(); + return 0; + default: + return -EINVAL; + } if (_IOC_DIR(cmd) & _IOC_READ) return fd_copyout((void __user *)param, outparam, size); - else - return 0; -#undef OUT -#undef IN + + return 0; } static void __init config_types(void) { - int first = 1; + bool has_drive = false; int drive; /* read drive info out of physical CMOS */ @@ -3655,17 +3627,22 @@ static void __init config_types(void) name = temparea; } if (name) { - const char *prepend = ","; - if (first) { - prepend = KERN_INFO "Floppy drive(s):"; - first = 0; + const char *prepend; + if (!has_drive) { + prepend = ""; + has_drive = true; + pr_info("Floppy drive(s):"); + } else { + prepend = ","; } - printk("%s fd%d is %s", prepend, drive, name); + + pr_cont("%s fd%d is %s", prepend, drive, name); } *UDP = *params; } - if (!first) - printk("\n"); + + if (has_drive) + pr_cont("\n"); } static int floppy_release(struct gendisk *disk, fmode_t mode) @@ -3705,8 +3682,8 @@ static int floppy_open(struct block_device *bdev, fmode_t mode) goto out2; if (!UDRS->fd_ref && (UDP->flags & FD_BROKEN_DCL)) { - USETF(FD_DISK_CHANGED); - USETF(FD_VERIFY); + set_bit(FD_DISK_CHANGED_BIT, &UDRS->flags); + set_bit(FD_VERIFY_BIT, &UDRS->flags); } if (UDRS->fd_ref == -1 || (UDRS->fd_ref && (mode & FMODE_EXCL))) @@ -3735,9 +3712,8 @@ static int floppy_open(struct block_device *bdev, fmode_t mode) INFBOUND(try, 16); tmp = (char *)fd_dma_mem_alloc(1024 * try); } - if (!tmp && !floppy_track_buffer) { + if (!tmp && !floppy_track_buffer) fallback_on_nodma_alloc(&tmp, 2048 * try); - } if (!tmp && !floppy_track_buffer) { DPRINT("Unable to allocate DMA memory\n"); goto out; @@ -3767,11 +3743,12 @@ static int floppy_open(struct block_device *bdev, fmode_t mode) if (mode & (FMODE_READ|FMODE_WRITE)) { UDRS->last_checked = 0; check_disk_change(bdev); - if (UTESTF(FD_DISK_CHANGED)) + if (test_bit(FD_DISK_CHANGED_BIT, &UDRS->flags)) goto out; } res = -EROFS; - if ((mode & FMODE_WRITE) && !(UTESTF(FD_DISK_WRITABLE))) + if ((mode & FMODE_WRITE) && + !test_bit(FD_DISK_WRITABLE_BIT, &UDRS->flags)) goto out; } mutex_unlock(&open_lock); @@ -3795,17 +3772,18 @@ static int check_floppy_change(struct gendisk *disk) { int drive = (long)disk->private_data; - if (UTESTF(FD_DISK_CHANGED) || UTESTF(FD_VERIFY)) + if (test_bit(FD_DISK_CHANGED_BIT, &UDRS->flags) || + test_bit(FD_VERIFY_BIT, &UDRS->flags)) return 1; if (time_after(jiffies, UDRS->last_checked + UDP->checkfreq)) { - lock_fdc(drive, 0); - poll_drive(0, 0); + lock_fdc(drive, false); + poll_drive(false, 0); process_fd_request(); } - if (UTESTF(FD_DISK_CHANGED) || - UTESTF(FD_VERIFY) || + if (test_bit(FD_DISK_CHANGED_BIT, &UDRS->flags) || + test_bit(FD_VERIFY_BIT, &UDRS->flags) || test_bit(drive, &fake_change) || (!ITYPE(UDRS->fd_device) && !current_type[drive])) return 1; @@ -3818,8 +3796,7 @@ static int check_floppy_change(struct gendisk *disk) * a disk in the drive, and whether that disk is writable. */ -static void floppy_rb0_complete(struct bio *bio, - int err) +static void floppy_rb0_complete(struct bio *bio, int err) { complete((struct completion *)bio->bi_private); } @@ -3877,14 +3854,16 @@ static int floppy_revalidate(struct gendisk *disk) int cf; int res = 0; - if (UTESTF(FD_DISK_CHANGED) || - UTESTF(FD_VERIFY) || test_bit(drive, &fake_change) || NO_GEOM) { + if (test_bit(FD_DISK_CHANGED_BIT, &UDRS->flags) || + test_bit(FD_VERIFY_BIT, &UDRS->flags) || + test_bit(drive, &fake_change) || NO_GEOM) { if (usage_count == 0) { - printk("VFS: revalidate called on non-open device.\n"); + pr_info("VFS: revalidate called on non-open device.\n"); return -EFAULT; } - lock_fdc(drive, 0); - cf = UTESTF(FD_DISK_CHANGED) || UTESTF(FD_VERIFY); + lock_fdc(drive, false); + cf = (test_bit(FD_DISK_CHANGED_BIT, &UDRS->flags) || + test_bit(FD_VERIFY_BIT, &UDRS->flags)); if (!(cf || test_bit(drive, &fake_change) || NO_GEOM)) { process_fd_request(); /*already done by another thread */ return 0; @@ -3894,7 +3873,7 @@ static int floppy_revalidate(struct gendisk *disk) if (buffer_drive == drive) buffer_track = -1; clear_bit(drive, &fake_change); - UCLEARF(FD_DISK_CHANGED); + clear_bit(FD_DISK_CHANGED_BIT, &UDRS->flags); if (cf) UDRS->generation++; if (NO_GEOM) { @@ -3902,7 +3881,7 @@ static int floppy_revalidate(struct gendisk *disk) res = __floppy_read_block_0(opened_bdev[drive]); } else { if (cf) - poll_drive(0, FD_RAW_NEED_DISK); + poll_drive(false, FD_RAW_NEED_DISK); process_fd_request(); } } @@ -3934,21 +3913,21 @@ static char __init get_fdc_version(void) output_byte(FD_DUMPREGS); /* 82072 and better know DUMPREGS */ if (FDCS->reset) return FDC_NONE; - if ((r = result()) <= 0x00) + r = result(); + if (r <= 0x00) return FDC_NONE; /* No FDC present ??? */ if ((r == 1) && (reply_buffer[0] == 0x80)) { - printk(KERN_INFO "FDC %d is an 8272A\n", fdc); + pr_info("FDC %d is an 8272A\n", fdc); return FDC_8272A; /* 8272a/765 don't know DUMPREGS */ } if (r != 10) { - printk - ("FDC %d init: DUMPREGS: unexpected return of %d bytes.\n", - fdc, r); + pr_info("FDC %d init: DUMPREGS: unexpected return of %d bytes.\n", + fdc, r); return FDC_UNKNOWN; } if (!fdc_configure()) { - printk(KERN_INFO "FDC %d is an 82072\n", fdc); + pr_info("FDC %d is an 82072\n", fdc); return FDC_82072; /* 82072 doesn't know CONFIGURE */ } @@ -3956,52 +3935,50 @@ static char __init get_fdc_version(void) if (need_more_output() == MORE_OUTPUT) { output_byte(0); } else { - printk(KERN_INFO "FDC %d is an 82072A\n", fdc); + pr_info("FDC %d is an 82072A\n", fdc); return FDC_82072A; /* 82072A as found on Sparcs. */ } output_byte(FD_UNLOCK); r = result(); if ((r == 1) && (reply_buffer[0] == 0x80)) { - printk(KERN_INFO "FDC %d is a pre-1991 82077\n", fdc); - return FDC_82077_ORIG; /* Pre-1991 82077, doesn't know + pr_info("FDC %d is a pre-1991 82077\n", fdc); + return FDC_82077_ORIG; /* Pre-1991 82077, doesn't know * LOCK/UNLOCK */ } if ((r != 1) || (reply_buffer[0] != 0x00)) { - printk("FDC %d init: UNLOCK: unexpected return of %d bytes.\n", - fdc, r); + pr_info("FDC %d init: UNLOCK: unexpected return of %d bytes.\n", + fdc, r); return FDC_UNKNOWN; } output_byte(FD_PARTID); r = result(); if (r != 1) { - printk("FDC %d init: PARTID: unexpected return of %d bytes.\n", - fdc, r); + pr_info("FDC %d init: PARTID: unexpected return of %d bytes.\n", + fdc, r); return FDC_UNKNOWN; } if (reply_buffer[0] == 0x80) { - printk(KERN_INFO "FDC %d is a post-1991 82077\n", fdc); + pr_info("FDC %d is a post-1991 82077\n", fdc); return FDC_82077; /* Revised 82077AA passes all the tests */ } switch (reply_buffer[0] >> 5) { case 0x0: /* Either a 82078-1 or a 82078SL running at 5Volt */ - printk(KERN_INFO "FDC %d is an 82078.\n", fdc); + pr_info("FDC %d is an 82078.\n", fdc); return FDC_82078; case 0x1: - printk(KERN_INFO "FDC %d is a 44pin 82078\n", fdc); + pr_info("FDC %d is a 44pin 82078\n", fdc); return FDC_82078; case 0x2: - printk(KERN_INFO "FDC %d is a S82078B\n", fdc); + pr_info("FDC %d is a S82078B\n", fdc); return FDC_S82078B; case 0x3: - printk(KERN_INFO "FDC %d is a National Semiconductor PC87306\n", - fdc); + pr_info("FDC %d is a National Semiconductor PC87306\n", fdc); return FDC_87306; default: - printk(KERN_INFO - "FDC %d init: 82078 variant with unknown PARTID=%d.\n", - fdc, reply_buffer[0] >> 5); + pr_info("FDC %d init: 82078 variant with unknown PARTID=%d.\n", + fdc, reply_buffer[0] >> 5); return FDC_82078_UNKN; } } /* get_fdc_version */ @@ -4113,9 +4090,9 @@ static int __init floppy_setup(char *str) else param = config_params[i].def_param; if (config_params[i].fn) - config_params[i]. - fn(ints, param, - config_params[i].param2); + config_params[i].fn(ints, param, + config_params[i]. + param2); if (config_params[i].var) { DPRINT("%s=%d\n", str, param); *config_params[i].var = param; @@ -4129,8 +4106,8 @@ static int __init floppy_setup(char *str) DPRINT("allowed options are:"); for (i = 0; i < ARRAY_SIZE(config_params); i++) - printk(" %s", config_params[i].name); - printk("\n"); + pr_cont(" %s", config_params[i].name); + pr_cont("\n"); } else DPRINT("botched floppy option\n"); DPRINT("Read Documentation/blockdev/floppy.txt\n"); @@ -4148,7 +4125,8 @@ static ssize_t floppy_cmos_show(struct device *dev, drive = p->id; return sprintf(buf, "%X\n", UDP->cmos); } -DEVICE_ATTR(cmos,S_IRUGO,floppy_cmos_show,NULL); + +DEVICE_ATTR(cmos, S_IRUGO, floppy_cmos_show, NULL); static void floppy_device_release(struct device *dev) { @@ -4160,7 +4138,7 @@ static int floppy_resume(struct device *dev) for (fdc = 0; fdc < N_FDC; fdc++) if (FDCS->address != -1) - user_reset_fdc(-1, FD_RESET_ALWAYS, 0); + user_reset_fdc(-1, FD_RESET_ALWAYS, false); return 0; } @@ -4172,8 +4150,8 @@ static const struct dev_pm_ops floppy_pm_ops = { static struct platform_driver floppy_driver = { .driver = { - .name = "floppy", - .pm = &floppy_pm_ops, + .name = "floppy", + .pm = &floppy_pm_ops, }, }; @@ -4234,7 +4212,7 @@ static int __init floppy_init(void) err = -ENOMEM; goto out_unreg_driver; } - blk_queue_max_sectors(floppy_queue, 64); + blk_queue_max_hw_sectors(floppy_queue, 64); blk_register_region(MKDEV(FLOPPY_MAJOR, 0), 256, THIS_MODULE, floppy_find, NULL, NULL); @@ -4245,16 +4223,16 @@ static int __init floppy_init(void) else floppy_sizes[i] = MAX_DISK_SIZE << 1; - reschedule_timeout(MAXTIMEOUT, "floppy init", MAXTIMEOUT); + reschedule_timeout(MAXTIMEOUT, "floppy init"); config_types(); for (i = 0; i < N_FDC; i++) { fdc = i; - CLEARSTRUCT(FDCS); + memset(FDCS, 0, sizeof(*FDCS)); FDCS->dtr = -1; FDCS->dor = 0x4; #if defined(__sparc__) || defined(__mc68000__) - /*sparcs/sun3x don't have a DOR reset which we can fall back on to */ + /*sparcs/sun3x don't have a DOR reset which we can fall back on to */ #ifdef __mc68000__ if (MACH_IS_SUN3X) #endif @@ -4283,11 +4261,11 @@ static int __init floppy_init(void) /* initialise drive state */ for (drive = 0; drive < N_DRIVE; drive++) { - CLEARSTRUCT(UDRS); - CLEARSTRUCT(UDRWE); - USETF(FD_DISK_NEWCHANGE); - USETF(FD_DISK_CHANGED); - USETF(FD_VERIFY); + memset(UDRS, 0, sizeof(*UDRS)); + memset(UDRWE, 0, sizeof(*UDRWE)); + set_bit(FD_DISK_NEWCHANGE_BIT, &UDRS->flags); + set_bit(FD_DISK_CHANGED_BIT, &UDRS->flags); + set_bit(FD_VERIFY_BIT, &UDRS->flags); UDRS->fd_device = -1; floppy_track_buffer = NULL; max_buffer_sectors = 0; @@ -4307,7 +4285,7 @@ static int __init floppy_init(void) if (FDCS->address == -1) continue; FDCS->rawcmd = 2; - if (user_reset_fdc(-1, FD_RESET_ALWAYS, 0)) { + if (user_reset_fdc(-1, FD_RESET_ALWAYS, false)) { /* free ioports reserved by floppy_grab_irq_and_dma() */ floppy_release_regions(fdc); FDCS->address = -1; @@ -4330,12 +4308,12 @@ static int __init floppy_init(void) * properly, so force a reset for the standard FDC clones, * to avoid interrupt garbage. */ - user_reset_fdc(-1, FD_RESET_ALWAYS, 0); + user_reset_fdc(-1, FD_RESET_ALWAYS, false); } fdc = 0; del_timer(&fd_timeout); current_drive = 0; - initialising = 0; + initialized = true; if (have_no_fdc) { DPRINT("no floppy controllers found\n"); err = have_no_fdc; @@ -4356,7 +4334,8 @@ static int __init floppy_init(void) if (err) goto out_flush_work; - err = device_create_file(&floppy_device[drive].dev,&dev_attr_cmos); + err = device_create_file(&floppy_device[drive].dev, + &dev_attr_cmos); if (err) goto out_unreg_platform_dev; @@ -4420,8 +4399,10 @@ static int floppy_request_regions(int fdc) const struct io_region *p; for (p = io_regions; p < ARRAY_END(io_regions); p++) { - if (!request_region(FDCS->address + p->offset, p->size, "floppy")) { - DPRINT("Floppy io-port 0x%04lx in use\n", FDCS->address + p->offset); + if (!request_region(FDCS->address + p->offset, + p->size, "floppy")) { + DPRINT("Floppy io-port 0x%04lx in use\n", + FDCS->address + p->offset); floppy_release_allocated_regions(fdc, p); return -EBUSY; } @@ -4512,11 +4493,9 @@ cleanup: static void floppy_release_irq_and_dma(void) { int old_fdc; -#ifdef FLOPPY_SANITY_CHECK #ifndef __sparc__ int drive; #endif -#endif long tmpsize; unsigned long tmpaddr; unsigned long flags; @@ -4547,20 +4526,18 @@ static void floppy_release_irq_and_dma(void) buffer_min = buffer_max = -1; fd_dma_mem_free(tmpaddr, tmpsize); } -#ifdef FLOPPY_SANITY_CHECK #ifndef __sparc__ for (drive = 0; drive < N_FDC * 4; drive++) if (timer_pending(motor_off_timer + drive)) - printk("motor off timer %d still active\n", drive); + pr_info("motor off timer %d still active\n", drive); #endif if (timer_pending(&fd_timeout)) - printk("floppy timer still active:%s\n", timeout_message); + pr_info("floppy timer still active:%s\n", timeout_message); if (timer_pending(&fd_timer)) - printk("auxiliary floppy timer still active\n"); + pr_info("auxiliary floppy timer still active\n"); if (work_pending(&floppy_work)) - printk("work still pending\n"); -#endif + pr_info("work still pending\n"); old_fdc = fdc; for (fdc = 0; fdc < N_FDC; fdc++) if (FDCS->address != -1) @@ -4577,7 +4554,9 @@ static void __init parse_floppy_cfg_string(char *cfg) char *ptr; while (*cfg) { - for (ptr = cfg; *cfg && *cfg != ' ' && *cfg != '\t'; cfg++) ; + ptr = cfg; + while (*cfg && *cfg != ' ' && *cfg != '\t') + cfg++; if (*cfg) { *cfg = '\0'; cfg++; @@ -4625,6 +4604,7 @@ static void __exit floppy_module_exit(void) /* eject disk, if any */ fd_eject(0); } + module_exit(floppy_module_exit); module_param(floppy, charp, 0); @@ -4636,9 +4616,10 @@ MODULE_LICENSE("GPL"); /* This doesn't actually get used other than for module information */ static const struct pnp_device_id floppy_pnpids[] = { - { "PNP0700", 0 }, - { } + {"PNP0700", 0}, + {} }; + MODULE_DEVICE_TABLE(pnp, floppy_pnpids); #else diff --git a/drivers/block/hd.c b/drivers/block/hd.c index d5cdce08ffd2..81c78b3ce2df 100644 --- a/drivers/block/hd.c +++ b/drivers/block/hd.c @@ -34,7 +34,6 @@ #include <linux/fs.h> #include <linux/kernel.h> #include <linux/genhd.h> -#include <linux/slab.h> #include <linux/string.h> #include <linux/ioport.h> #include <linux/init.h> @@ -165,12 +164,12 @@ unsigned long read_timer(void) unsigned long t, flags; int i; - spin_lock_irqsave(&i8253_lock, flags); + raw_spin_lock_irqsave(&i8253_lock, flags); t = jiffies * 11932; outb_p(0, 0x43); i = inb_p(0x40); i |= inb(0x40) << 8; - spin_unlock_irqrestore(&i8253_lock, flags); + raw_spin_unlock_irqrestore(&i8253_lock, flags); return(t - i); } #endif @@ -719,7 +718,7 @@ static int __init hd_init(void) return -ENOMEM; } - blk_queue_max_sectors(hd_queue, 255); + blk_queue_max_hw_sectors(hd_queue, 255); init_timer(&device_timer); device_timer.function = hd_times_out; blk_queue_logical_block_size(hd_queue, 512); diff --git a/drivers/block/loop.c b/drivers/block/loop.c index bd112c8c7bcd..6120922f459f 100644 --- a/drivers/block/loop.c +++ b/drivers/block/loop.c @@ -71,7 +71,6 @@ #include <linux/buffer_head.h> /* for invalidate_bdev() */ #include <linux/completion.h> #include <linux/highmem.h> -#include <linux/gfp.h> #include <linux/kthread.h> #include <linux/splice.h> @@ -238,6 +237,8 @@ static int do_lo_send_aops(struct loop_device *lo, struct bio_vec *bvec, if (ret) goto fail; + file_update_time(file); + transfer_result = lo_do_transfer(lo, WRITE, page, offset, bvec->bv_page, bv_offs, size, IV); copied = size; @@ -484,7 +485,7 @@ static int do_bio_filebacked(struct loop_device *lo, struct bio *bio) goto out; } - ret = vfs_fsync(file, file->f_path.dentry, 0); + ret = vfs_fsync(file, 0); if (unlikely(ret)) { ret = -EIO; goto out; @@ -494,7 +495,7 @@ static int do_bio_filebacked(struct loop_device *lo, struct bio *bio) ret = lo_send(lo, bio, pos); if (barrier && !ret) { - ret = vfs_fsync(file, file->f_path.dentry, 0); + ret = vfs_fsync(file, 0); if (unlikely(ret)) ret = -EIO; } @@ -834,6 +835,8 @@ static int loop_set_fd(struct loop_device *lo, fmode_t mode, set_capacity(lo->lo_disk, size); bd_set_size(bdev, size << 9); + /* let user-space know about the new size */ + kobject_uevent(&disk_to_dev(bdev->bd_disk)->kobj, KOBJ_CHANGE); set_blocksize(bdev, lo_blocksize); @@ -857,6 +860,7 @@ out_clr: set_capacity(lo->lo_disk, 0); invalidate_bdev(bdev); bd_set_size(bdev, 0); + kobject_uevent(&disk_to_dev(bdev->bd_disk)->kobj, KOBJ_CHANGE); mapping_set_gfp_mask(mapping, lo->old_gfp_mask); lo->lo_state = Lo_unbound; out_putf: @@ -943,8 +947,11 @@ static int loop_clr_fd(struct loop_device *lo, struct block_device *bdev) if (bdev) invalidate_bdev(bdev); set_capacity(lo->lo_disk, 0); - if (bdev) + if (bdev) { bd_set_size(bdev, 0); + /* let user-space know about this change */ + kobject_uevent(&disk_to_dev(bdev->bd_disk)->kobj, KOBJ_CHANGE); + } mapping_set_gfp_mask(filp->f_mapping, gfp); lo->lo_state = Lo_unbound; /* This is safe: open() is still holding a reference. */ @@ -1188,6 +1195,8 @@ static int loop_set_capacity(struct loop_device *lo, struct block_device *bdev) sz <<= 9; mutex_lock(&bdev->bd_mutex); bd_set_size(bdev, sz); + /* let user-space know about the new size */ + kobject_uevent(&disk_to_dev(bdev->bd_disk)->kobj, KOBJ_CHANGE); mutex_unlock(&bdev->bd_mutex); out: diff --git a/drivers/block/mg_disk.c b/drivers/block/mg_disk.c index 02b2583df7fc..28db925dbdad 100644 --- a/drivers/block/mg_disk.c +++ b/drivers/block/mg_disk.c @@ -23,6 +23,7 @@ #include <linux/platform_device.h> #include <linux/gpio.h> #include <linux/mg_disk.h> +#include <linux/slab.h> #define MG_RES_SEC (CONFIG_MG_DISK_RES << 1) @@ -980,7 +981,7 @@ static int mg_probe(struct platform_device *plat_dev) __func__, __LINE__); goto probe_err_6; } - blk_queue_max_sectors(host->breq, MG_MAX_SECTS); + blk_queue_max_hw_sectors(host->breq, MG_MAX_SECTS); blk_queue_logical_block_size(host->breq, MG_SECTOR_SIZE); init_timer(&host->timer); diff --git a/drivers/block/nbd.c b/drivers/block/nbd.c index cc923a5b430c..218d091f3c52 100644 --- a/drivers/block/nbd.c +++ b/drivers/block/nbd.c @@ -27,6 +27,7 @@ #include <linux/compiler.h> #include <linux/err.h> #include <linux/kernel.h> +#include <linux/slab.h> #include <net/sock.h> #include <linux/net.h> #include <linux/kthread.h> diff --git a/drivers/block/osdblk.c b/drivers/block/osdblk.c index a808b1530b3b..6cd8b705b11b 100644 --- a/drivers/block/osdblk.c +++ b/drivers/block/osdblk.c @@ -63,6 +63,7 @@ #include <linux/device.h> #include <linux/module.h> #include <linux/fs.h> +#include <linux/slab.h> #include <scsi/osd_initiator.h> #include <scsi/osd_attributes.h> #include <scsi/osd_sec.h> @@ -476,7 +477,9 @@ static void class_osdblk_release(struct class *cls) kfree(cls); } -static ssize_t class_osdblk_list(struct class *c, char *data) +static ssize_t class_osdblk_list(struct class *c, + struct class_attribute *attr, + char *data) { int n = 0; struct list_head *tmp; @@ -500,7 +503,9 @@ static ssize_t class_osdblk_list(struct class *c, char *data) return n; } -static ssize_t class_osdblk_add(struct class *c, const char *buf, size_t count) +static ssize_t class_osdblk_add(struct class *c, + struct class_attribute *attr, + const char *buf, size_t count) { struct osdblk_device *osdev; ssize_t rc; @@ -592,7 +597,9 @@ err_out_mod: return rc; } -static ssize_t class_osdblk_remove(struct class *c, const char *buf, +static ssize_t class_osdblk_remove(struct class *c, + struct class_attribute *attr, + const char *buf, size_t count) { struct osdblk_device *osdev = NULL; diff --git a/drivers/block/paride/pcd.c b/drivers/block/paride/pcd.c index 8866ca369d5e..71acf4e53356 100644 --- a/drivers/block/paride/pcd.c +++ b/drivers/block/paride/pcd.c @@ -341,11 +341,11 @@ static int pcd_wait(struct pcd_unit *cd, int go, int stop, char *fun, char *msg) && (j++ < PCD_SPIN)) udelay(PCD_DELAY); - if ((r & (IDE_ERR & stop)) || (j >= PCD_SPIN)) { + if ((r & (IDE_ERR & stop)) || (j > PCD_SPIN)) { s = read_reg(cd, 7); e = read_reg(cd, 1); p = read_reg(cd, 2); - if (j >= PCD_SPIN) + if (j > PCD_SPIN) e |= 0x100; if (fun) printk("%s: %s %s: alt=0x%x stat=0x%x err=0x%x" diff --git a/drivers/block/paride/pd.c b/drivers/block/paride/pd.c index 569e39e8f114..c1e5cd029b23 100644 --- a/drivers/block/paride/pd.c +++ b/drivers/block/paride/pd.c @@ -145,6 +145,7 @@ enum {D_PRT, D_PRO, D_UNI, D_MOD, D_GEO, D_SBY, D_DLY, D_SLV}; #include <linux/init.h> #include <linux/module.h> +#include <linux/gfp.h> #include <linux/fs.h> #include <linux/delay.h> #include <linux/hdreg.h> @@ -906,7 +907,7 @@ static int __init pd_init(void) if (!pd_queue) goto out1; - blk_queue_max_sectors(pd_queue, cluster); + blk_queue_max_hw_sectors(pd_queue, cluster); if (register_blkdev(major, name)) goto out2; diff --git a/drivers/block/paride/pf.c b/drivers/block/paride/pf.c index ea54ea393553..c059aab3006b 100644 --- a/drivers/block/paride/pf.c +++ b/drivers/block/paride/pf.c @@ -391,11 +391,11 @@ static int pf_wait(struct pf_unit *pf, int go, int stop, char *fun, char *msg) && (j++ < PF_SPIN)) udelay(PF_SPIN_DEL); - if ((r & (STAT_ERR & stop)) || (j >= PF_SPIN)) { + if ((r & (STAT_ERR & stop)) || (j > PF_SPIN)) { s = read_reg(pf, 7); e = read_reg(pf, 1); p = read_reg(pf, 2); - if (j >= PF_SPIN) + if (j > PF_SPIN) e |= 0x100; if (fun) printk("%s: %s %s: alt=0x%x stat=0x%x err=0x%x" @@ -956,8 +956,7 @@ static int __init pf_init(void) return -ENOMEM; } - blk_queue_max_phys_segments(pf_queue, cluster); - blk_queue_max_hw_segments(pf_queue, cluster); + blk_queue_max_segments(pf_queue, cluster); for (pf = units, unit = 0; unit < PF_UNITS; pf++, unit++) { struct gendisk *disk = pf->disk; diff --git a/drivers/block/paride/pt.c b/drivers/block/paride/pt.c index 1e4006e18f03..bc5825fdeaab 100644 --- a/drivers/block/paride/pt.c +++ b/drivers/block/paride/pt.c @@ -274,11 +274,11 @@ static int pt_wait(struct pt_unit *tape, int go, int stop, char *fun, char *msg) && (j++ < PT_SPIN)) udelay(PT_SPIN_DEL); - if ((r & (STAT_ERR & stop)) || (j >= PT_SPIN)) { + if ((r & (STAT_ERR & stop)) || (j > PT_SPIN)) { s = read_reg(pi, 7); e = read_reg(pi, 1); p = read_reg(pi, 2); - if (j >= PT_SPIN) + if (j > PT_SPIN) e |= 0x100; if (fun) printk("%s: %s %s: alt=0x%x stat=0x%x err=0x%x" diff --git a/drivers/block/pktcdvd.c b/drivers/block/pktcdvd.c index 68b5957f107c..8a549db2aa78 100644 --- a/drivers/block/pktcdvd.c +++ b/drivers/block/pktcdvd.c @@ -48,6 +48,7 @@ #include <linux/module.h> #include <linux/types.h> #include <linux/kernel.h> +#include <linux/compat.h> #include <linux/kthread.h> #include <linux/errno.h> #include <linux/spinlock.h> @@ -57,6 +58,7 @@ #include <linux/miscdevice.h> #include <linux/freezer.h> #include <linux/mutex.h> +#include <linux/slab.h> #include <scsi/scsi_cmnd.h> #include <scsi/scsi_ioctl.h> #include <scsi/scsi.h> @@ -284,7 +286,7 @@ static ssize_t kobj_pkt_store(struct kobject *kobj, return len; } -static struct sysfs_ops kobj_pkt_ops = { +static const struct sysfs_ops kobj_pkt_ops = { .show = kobj_pkt_show, .store = kobj_pkt_store }; @@ -337,7 +339,9 @@ static void class_pktcdvd_release(struct class *cls) { kfree(cls); } -static ssize_t class_pktcdvd_show_map(struct class *c, char *data) +static ssize_t class_pktcdvd_show_map(struct class *c, + struct class_attribute *attr, + char *data) { int n = 0; int idx; @@ -356,7 +360,9 @@ static ssize_t class_pktcdvd_show_map(struct class *c, char *data) return n; } -static ssize_t class_pktcdvd_store_add(struct class *c, const char *buf, +static ssize_t class_pktcdvd_store_add(struct class *c, + struct class_attribute *attr, + const char *buf, size_t count) { unsigned int major, minor; @@ -376,7 +382,9 @@ static ssize_t class_pktcdvd_store_add(struct class *c, const char *buf, return -EINVAL; } -static ssize_t class_pktcdvd_store_remove(struct class *c, const char *buf, +static ssize_t class_pktcdvd_store_remove(struct class *c, + struct class_attribute *attr, + const char *buf, size_t count) { unsigned int major, minor; @@ -569,6 +577,7 @@ static struct packet_data *pkt_alloc_packet_data(int frames) } spin_lock_init(&pkt->lock); + bio_list_init(&pkt->orig_bios); for (i = 0; i < frames; i++) { struct bio *bio = pkt_bio_alloc(1); @@ -721,43 +730,6 @@ static void pkt_rbtree_insert(struct pktcdvd_device *pd, struct pkt_rb_node *nod } /* - * Add a bio to a single linked list defined by its head and tail pointers. - */ -static void pkt_add_list_last(struct bio *bio, struct bio **list_head, struct bio **list_tail) -{ - bio->bi_next = NULL; - if (*list_tail) { - BUG_ON((*list_head) == NULL); - (*list_tail)->bi_next = bio; - (*list_tail) = bio; - } else { - BUG_ON((*list_head) != NULL); - (*list_head) = bio; - (*list_tail) = bio; - } -} - -/* - * Remove and return the first bio from a single linked list defined by its - * head and tail pointers. - */ -static inline struct bio *pkt_get_list_first(struct bio **list_head, struct bio **list_tail) -{ - struct bio *bio; - - if (*list_head == NULL) - return NULL; - - bio = *list_head; - *list_head = bio->bi_next; - if (*list_head == NULL) - *list_tail = NULL; - - bio->bi_next = NULL; - return bio; -} - -/* * Send a packet_command to the underlying block device and * wait for completion. */ @@ -876,13 +848,10 @@ static noinline_for_stack int pkt_set_speed(struct pktcdvd_device *pd, static void pkt_queue_bio(struct pktcdvd_device *pd, struct bio *bio) { spin_lock(&pd->iosched.lock); - if (bio_data_dir(bio) == READ) { - pkt_add_list_last(bio, &pd->iosched.read_queue, - &pd->iosched.read_queue_tail); - } else { - pkt_add_list_last(bio, &pd->iosched.write_queue, - &pd->iosched.write_queue_tail); - } + if (bio_data_dir(bio) == READ) + bio_list_add(&pd->iosched.read_queue, bio); + else + bio_list_add(&pd->iosched.write_queue, bio); spin_unlock(&pd->iosched.lock); atomic_set(&pd->iosched.attention, 1); @@ -917,8 +886,8 @@ static void pkt_iosched_process_queue(struct pktcdvd_device *pd) int reads_queued, writes_queued; spin_lock(&pd->iosched.lock); - reads_queued = (pd->iosched.read_queue != NULL); - writes_queued = (pd->iosched.write_queue != NULL); + reads_queued = !bio_list_empty(&pd->iosched.read_queue); + writes_queued = !bio_list_empty(&pd->iosched.write_queue); spin_unlock(&pd->iosched.lock); if (!reads_queued && !writes_queued) @@ -927,7 +896,7 @@ static void pkt_iosched_process_queue(struct pktcdvd_device *pd) if (pd->iosched.writing) { int need_write_seek = 1; spin_lock(&pd->iosched.lock); - bio = pd->iosched.write_queue; + bio = bio_list_peek(&pd->iosched.write_queue); spin_unlock(&pd->iosched.lock); if (bio && (bio->bi_sector == pd->iosched.last_write)) need_write_seek = 0; @@ -950,13 +919,10 @@ static void pkt_iosched_process_queue(struct pktcdvd_device *pd) } spin_lock(&pd->iosched.lock); - if (pd->iosched.writing) { - bio = pkt_get_list_first(&pd->iosched.write_queue, - &pd->iosched.write_queue_tail); - } else { - bio = pkt_get_list_first(&pd->iosched.read_queue, - &pd->iosched.read_queue_tail); - } + if (pd->iosched.writing) + bio = bio_list_pop(&pd->iosched.write_queue); + else + bio = bio_list_pop(&pd->iosched.read_queue); spin_unlock(&pd->iosched.lock); if (!bio) @@ -992,14 +958,14 @@ static void pkt_iosched_process_queue(struct pktcdvd_device *pd) static int pkt_set_segment_merging(struct pktcdvd_device *pd, struct request_queue *q) { if ((pd->settings.size << 9) / CD_FRAMESIZE - <= queue_max_phys_segments(q)) { + <= queue_max_segments(q)) { /* * The cdrom device can handle one segment/frame */ clear_bit(PACKET_MERGE_SEGS, &pd->flags); return 0; } else if ((pd->settings.size << 9) / PAGE_SIZE - <= queue_max_phys_segments(q)) { + <= queue_max_segments(q)) { /* * We can handle this case at the expense of some extra memory * copies during write operations @@ -1114,7 +1080,7 @@ static void pkt_gather_data(struct pktcdvd_device *pd, struct packet_data *pkt) int f; char written[PACKET_MAX_SIZE]; - BUG_ON(!pkt->orig_bios); + BUG_ON(bio_list_empty(&pkt->orig_bios)); atomic_set(&pkt->io_wait, 0); atomic_set(&pkt->io_errors, 0); @@ -1124,7 +1090,7 @@ static void pkt_gather_data(struct pktcdvd_device *pd, struct packet_data *pkt) */ memset(written, 0, sizeof(written)); spin_lock(&pkt->lock); - for (bio = pkt->orig_bios; bio; bio = bio->bi_next) { + bio_list_for_each(bio, &pkt->orig_bios) { int first_frame = (bio->bi_sector - pkt->sector) / (CD_FRAMESIZE >> 9); int num_frames = bio->bi_size / CD_FRAMESIZE; pd->stats.secs_w += num_frames * (CD_FRAMESIZE >> 9); @@ -1363,7 +1329,7 @@ try_next_bio: break; pkt_rbtree_erase(pd, node); spin_lock(&pkt->lock); - pkt_add_list_last(bio, &pkt->orig_bios, &pkt->orig_bios_tail); + bio_list_add(&pkt->orig_bios, bio); pkt->write_size += bio->bi_size / CD_FRAMESIZE; spin_unlock(&pkt->lock); } @@ -1409,7 +1375,7 @@ static void pkt_start_write(struct pktcdvd_device *pd, struct packet_data *pkt) */ frames_write = 0; spin_lock(&pkt->lock); - for (bio = pkt->orig_bios; bio; bio = bio->bi_next) { + bio_list_for_each(bio, &pkt->orig_bios) { int segment = bio->bi_idx; int src_offs = 0; int first_frame = (bio->bi_sector - pkt->sector) / (CD_FRAMESIZE >> 9); @@ -1472,20 +1438,14 @@ static void pkt_start_write(struct pktcdvd_device *pd, struct packet_data *pkt) static void pkt_finish_packet(struct packet_data *pkt, int uptodate) { - struct bio *bio, *next; + struct bio *bio; if (!uptodate) pkt->cache_valid = 0; /* Finish all bios corresponding to this packet */ - bio = pkt->orig_bios; - while (bio) { - next = bio->bi_next; - bio->bi_next = NULL; + while ((bio = bio_list_pop(&pkt->orig_bios))) bio_endio(bio, uptodate ? 0 : -EIO); - bio = next; - } - pkt->orig_bios = pkt->orig_bios_tail = NULL; } static void pkt_run_state_machine(struct pktcdvd_device *pd, struct packet_data *pkt) @@ -2360,7 +2320,7 @@ static int pkt_open_dev(struct pktcdvd_device *pd, fmode_t write) * even if the size is a multiple of the packet size. */ spin_lock_irq(q->queue_lock); - blk_queue_max_sectors(q, pd->settings.size); + blk_queue_max_hw_sectors(q, pd->settings.size); spin_unlock_irq(q->queue_lock); set_bit(PACKET_WRITABLE, &pd->flags); } else { @@ -2567,8 +2527,7 @@ static int pkt_make_request(struct request_queue *q, struct bio *bio) spin_lock(&pkt->lock); if ((pkt->state == PACKET_WAITING_STATE) || (pkt->state == PACKET_READ_WAIT_STATE)) { - pkt_add_list_last(bio, &pkt->orig_bios, - &pkt->orig_bios_tail); + bio_list_add(&pkt->orig_bios, bio); pkt->write_size += bio->bi_size / CD_FRAMESIZE; if ((pkt->write_size >= pkt->frames) && (pkt->state == PACKET_WAITING_STATE)) { @@ -2662,7 +2621,7 @@ static void pkt_init_queue(struct pktcdvd_device *pd) blk_queue_make_request(q, pkt_make_request); blk_queue_logical_block_size(q, CD_FRAMESIZE); - blk_queue_max_sectors(q, PACKET_MAX_SECTORS); + blk_queue_max_hw_sectors(q, PACKET_MAX_SECTORS); blk_queue_merge_bvec(q, pkt_merge_bvec); q->queuedata = pd; } @@ -2898,6 +2857,8 @@ static int pkt_setup_dev(dev_t dev, dev_t* pkt_dev) spin_lock_init(&pd->lock); spin_lock_init(&pd->iosched.lock); + bio_list_init(&pd->iosched.read_queue); + bio_list_init(&pd->iosched.write_queue); sprintf(pd->name, DRIVER_NAME"%d", idx); init_waitqueue_head(&pd->wqueue); pd->bio_queue = RB_ROOT; @@ -3024,7 +2985,7 @@ static void pkt_get_status(struct pkt_ctrl_command *ctrl_cmd) mutex_unlock(&ctl_mutex); } -static int pkt_ctl_ioctl(struct inode *inode, struct file *file, unsigned int cmd, unsigned long arg) +static long pkt_ctl_ioctl(struct file *file, unsigned int cmd, unsigned long arg) { void __user *argp = (void __user *)arg; struct pkt_ctrl_command ctrl_cmd; @@ -3061,10 +3022,20 @@ static int pkt_ctl_ioctl(struct inode *inode, struct file *file, unsigned int cm return ret; } +#ifdef CONFIG_COMPAT +static long pkt_ctl_compat_ioctl(struct file *file, unsigned int cmd, unsigned long arg) +{ + return pkt_ctl_ioctl(file, cmd, (unsigned long)compat_ptr(arg)); +} +#endif static const struct file_operations pkt_ctl_fops = { - .ioctl = pkt_ctl_ioctl, - .owner = THIS_MODULE, + .open = nonseekable_open, + .unlocked_ioctl = pkt_ctl_ioctl, +#ifdef CONFIG_COMPAT + .compat_ioctl = pkt_ctl_compat_ioctl, +#endif + .owner = THIS_MODULE, }; static struct miscdevice pkt_misc = { diff --git a/drivers/block/ps3disk.c b/drivers/block/ps3disk.c index 03a130dca8ab..3b419e3fffa1 100644 --- a/drivers/block/ps3disk.c +++ b/drivers/block/ps3disk.c @@ -20,6 +20,7 @@ #include <linux/ata.h> #include <linux/blkdev.h> +#include <linux/slab.h> #include <asm/lv1call.h> #include <asm/ps3stor.h> @@ -474,7 +475,7 @@ static int __devinit ps3disk_probe(struct ps3_system_bus_device *_dev) blk_queue_bounce_limit(queue, BLK_BOUNCE_HIGH); - blk_queue_max_sectors(queue, dev->bounce_size >> 9); + blk_queue_max_hw_sectors(queue, dev->bounce_size >> 9); blk_queue_segment_boundary(queue, -1UL); blk_queue_dma_alignment(queue, dev->blk_size-1); blk_queue_logical_block_size(queue, dev->blk_size); @@ -482,8 +483,7 @@ static int __devinit ps3disk_probe(struct ps3_system_bus_device *_dev) blk_queue_ordered(queue, QUEUE_ORDERED_DRAIN_FLUSH, ps3disk_prepare_flush); - blk_queue_max_phys_segments(queue, -1); - blk_queue_max_hw_segments(queue, -1); + blk_queue_max_segments(queue, -1); blk_queue_max_segment_size(queue, dev->bounce_size); gendisk = alloc_disk(PS3DISK_MINORS); diff --git a/drivers/block/ps3vram.c b/drivers/block/ps3vram.c index 1fb6c3135fc8..b3bdb8af89cf 100644 --- a/drivers/block/ps3vram.c +++ b/drivers/block/ps3vram.c @@ -12,6 +12,7 @@ #include <linux/delay.h> #include <linux/proc_fs.h> #include <linux/seq_file.h> +#include <linux/slab.h> #include <asm/cell-regs.h> #include <asm/firmware.h> @@ -751,10 +752,9 @@ static int __devinit ps3vram_probe(struct ps3_system_bus_device *dev) priv->queue = queue; queue->queuedata = dev; blk_queue_make_request(queue, ps3vram_make_request); - blk_queue_max_phys_segments(queue, MAX_PHYS_SEGMENTS); - blk_queue_max_hw_segments(queue, MAX_HW_SEGMENTS); - blk_queue_max_segment_size(queue, MAX_SEGMENT_SIZE); - blk_queue_max_sectors(queue, SAFE_MAX_SECTORS); + blk_queue_max_segments(queue, BLK_MAX_SEGMENTS); + blk_queue_max_segment_size(queue, BLK_MAX_SEGMENT_SIZE); + blk_queue_max_hw_sectors(queue, BLK_SAFE_MAX_SECTORS); gendisk = alloc_disk(1); if (!gendisk) { diff --git a/drivers/block/sunvdc.c b/drivers/block/sunvdc.c index 411f064760b4..48e8fee9f2d4 100644 --- a/drivers/block/sunvdc.c +++ b/drivers/block/sunvdc.c @@ -691,9 +691,8 @@ static int probe_disk(struct vdc_port *port) port->disk = g; - blk_queue_max_hw_segments(q, port->ring_cookies); - blk_queue_max_phys_segments(q, port->ring_cookies); - blk_queue_max_sectors(q, port->max_xfer_size); + blk_queue_max_segments(q, port->ring_cookies); + blk_queue_max_hw_sectors(q, port->max_xfer_size); g->major = vdc_major; g->first_minor = port->vio.vdev->dev_no << PARTITION_SHIFT; strcpy(g->disk_name, port->disk_name); diff --git a/drivers/block/swim.c b/drivers/block/swim.c index 8f569e3df890..e463657569ff 100644 --- a/drivers/block/swim.c +++ b/drivers/block/swim.c @@ -18,6 +18,7 @@ #include <linux/module.h> #include <linux/fd.h> +#include <linux/slab.h> #include <linux/blkdev.h> #include <linux/hdreg.h> #include <linux/kernel.h> @@ -864,7 +865,7 @@ static int __devinit swim_probe(struct platform_device *dev) struct swim_priv *swd; int ret; - res = platform_get_resource_byname(dev, IORESOURCE_MEM, "swim-regs"); + res = platform_get_resource(dev, IORESOURCE_MEM, 0); if (!res) { ret = -ENODEV; goto out; @@ -942,7 +943,7 @@ static int __devexit swim_remove(struct platform_device *dev) iounmap(swd->base); - res = platform_get_resource_byname(dev, IORESOURCE_MEM, "swim-regs"); + res = platform_get_resource(dev, IORESOURCE_MEM, 0); if (res) release_mem_region(res->start, resource_size(res)); diff --git a/drivers/block/swim3.c b/drivers/block/swim3.c index 59ca2b77b574..ed6fb91123ab 100644 --- a/drivers/block/swim3.c +++ b/drivers/block/swim3.c @@ -1004,7 +1004,7 @@ static const struct block_device_operations floppy_fops = { static int swim3_add_device(struct macio_dev *mdev, int index) { - struct device_node *swim = mdev->ofdev.node; + struct device_node *swim = mdev->ofdev.dev.of_node; struct floppy_state *fs = &floppy_states[index]; int rc = -EBUSY; @@ -1159,8 +1159,10 @@ static struct of_device_id swim3_match[] = static struct macio_driver swim3_driver = { - .name = "swim3", - .match_table = swim3_match, + .driver = { + .name = "swim3", + .of_match_table = swim3_match, + }, .probe = swim3_attach, #if 0 .suspend = swim3_suspend, diff --git a/drivers/block/sx8.c b/drivers/block/sx8.c index a7c4184f4a63..b70f0fca9a42 100644 --- a/drivers/block/sx8.c +++ b/drivers/block/sx8.c @@ -409,7 +409,7 @@ static int carm_init_one (struct pci_dev *pdev, const struct pci_device_id *ent) static void carm_remove_one (struct pci_dev *pdev); static int carm_bdev_getgeo(struct block_device *bdev, struct hd_geometry *geo); -static struct pci_device_id carm_pci_tbl[] = { +static const struct pci_device_id carm_pci_tbl[] = { { PCI_VENDOR_ID_PROMISE, 0x8000, PCI_ANY_ID, PCI_ANY_ID, 0, 0, }, { PCI_VENDOR_ID_PROMISE, 0x8002, PCI_ANY_ID, PCI_ANY_ID, 0, 0, }, { } /* terminate list */ @@ -1518,8 +1518,7 @@ static int carm_init_disks(struct carm_host *host) break; } disk->queue = q; - blk_queue_max_hw_segments(q, CARM_MAX_REQ_SG); - blk_queue_max_phys_segments(q, CARM_MAX_REQ_SG); + blk_queue_max_segments(q, CARM_MAX_REQ_SG); blk_queue_segment_boundary(q, CARM_SG_BOUNDARY); q->queuedata = port; diff --git a/drivers/block/ub.c b/drivers/block/ub.c index c739b203fe91..0536b5b29adc 100644 --- a/drivers/block/ub.c +++ b/drivers/block/ub.c @@ -27,6 +27,7 @@ #include <linux/blkdev.h> #include <linux/timer.h> #include <linux/scatterlist.h> +#include <linux/slab.h> #include <scsi/scsi.h> #define DRV_NAME "ub" @@ -393,7 +394,7 @@ static int ub_probe_lun(struct ub_dev *sc, int lnum); #define ub_usb_ids usb_storage_usb_ids #else -static struct usb_device_id ub_usb_ids[] = { +static const struct usb_device_id ub_usb_ids[] = { { USB_INTERFACE_INFO(USB_CLASS_MASS_STORAGE, US_SC_SCSI, US_PR_BULK) }, { } }; @@ -2320,10 +2321,9 @@ static int ub_probe_lun(struct ub_dev *sc, int lnum) disk->queue = q; blk_queue_bounce_limit(q, BLK_BOUNCE_HIGH); - blk_queue_max_hw_segments(q, UB_MAX_REQ_SG); - blk_queue_max_phys_segments(q, UB_MAX_REQ_SG); + blk_queue_max_segments(q, UB_MAX_REQ_SG); blk_queue_segment_boundary(q, 0xffffffff); /* Dubious. */ - blk_queue_max_sectors(q, UB_MAX_SECTORS); + blk_queue_max_hw_sectors(q, UB_MAX_SECTORS); blk_queue_logical_block_size(q, lun->capacity.bsize); lun->disk = disk; diff --git a/drivers/block/umem.c b/drivers/block/umem.c index ad1ba393801a..2f9470ff8f7c 100644 --- a/drivers/block/umem.c +++ b/drivers/block/umem.c @@ -40,13 +40,13 @@ #include <linux/kernel.h> #include <linux/mm.h> #include <linux/mman.h> +#include <linux/gfp.h> #include <linux/ioctl.h> #include <linux/module.h> #include <linux/init.h> #include <linux/interrupt.h> #include <linux/timer.h> #include <linux/pci.h> -#include <linux/slab.h> #include <linux/dma-mapping.h> #include <linux/fcntl.h> /* O_ACCMODE */ diff --git a/drivers/block/viodasd.c b/drivers/block/viodasd.c index a8c8b56b275e..788d93882ab9 100644 --- a/drivers/block/viodasd.c +++ b/drivers/block/viodasd.c @@ -28,6 +28,9 @@ * All disk operations are performed by sending messages back and forth to * the OS/400 partition. */ + +#define pr_fmt(fmt) "viod: " fmt + #include <linux/major.h> #include <linux/fs.h> #include <linux/module.h> @@ -63,9 +66,6 @@ MODULE_LICENSE("GPL"); #define VIOD_VERS "1.64" -#define VIOD_KERN_WARNING KERN_WARNING "viod: " -#define VIOD_KERN_INFO KERN_INFO "viod: " - enum { PARTITION_SHIFT = 3, MAX_DISKNO = HVMAXARCHITECTEDVIRTUALDISKS, @@ -156,7 +156,7 @@ static int viodasd_open(struct block_device *bdev, fmode_t mode) ((u64)DEVICE_NO(d) << 48) | ((u64)flags << 32), 0, 0, 0); if (hvrc != 0) { - printk(VIOD_KERN_WARNING "HV open failed %d\n", (int)hvrc); + pr_warning("HV open failed %d\n", (int)hvrc); return -EIO; } @@ -167,9 +167,8 @@ static int viodasd_open(struct block_device *bdev, fmode_t mode) const struct vio_error_entry *err = vio_lookup_rc(viodasd_err_table, we.sub_result); - printk(VIOD_KERN_WARNING - "bad rc opening disk: %d:0x%04x (%s)\n", - (int)we.rc, we.sub_result, err->msg); + pr_warning("bad rc opening disk: %d:0x%04x (%s)\n", + (int)we.rc, we.sub_result, err->msg); return -EIO; } @@ -195,8 +194,7 @@ static int viodasd_release(struct gendisk *disk, fmode_t mode) ((u64)DEVICE_NO(d) << 48) /* | ((u64)flags << 32) */, 0, 0, 0); if (hvrc != 0) - printk(VIOD_KERN_WARNING "HV close call failed %d\n", - (int)hvrc); + pr_warning("HV close call failed %d\n", (int)hvrc); return 0; } @@ -288,8 +286,7 @@ static int send_request(struct request *req) bevent = (struct vioblocklpevent *) vio_get_event_buffer(viomajorsubtype_blockio); if (bevent == NULL) { - printk(VIOD_KERN_WARNING - "error allocating disk event buffer\n"); + pr_warning("error allocating disk event buffer\n"); goto error_ret; } @@ -333,9 +330,8 @@ static int send_request(struct request *req) } if (hvrc != HvLpEvent_Rc_Good) { - printk(VIOD_KERN_WARNING - "error sending disk event to OS/400 (rc %d)\n", - (int)hvrc); + pr_warning("error sending disk event to OS/400 (rc %d)\n", + (int)hvrc); goto error_ret; } spin_unlock_irqrestore(&viodasd_spinlock, flags); @@ -402,7 +398,7 @@ retry: ((u64)dev_no << 48) | ((u64)flags<< 32), 0, 0, 0); if (hvrc != 0) { - printk(VIOD_KERN_WARNING "bad rc on HV open %d\n", (int)hvrc); + pr_warning("bad rc on HV open %d\n", (int)hvrc); return 0; } @@ -416,9 +412,8 @@ retry: goto retry; } if (we.max_disk > (MAX_DISKNO - 1)) { - printk_once(VIOD_KERN_INFO - "Only examining the first %d of %d disks connected\n", - MAX_DISKNO, we.max_disk + 1); + printk_once(KERN_INFO pr_fmt("Only examining the first %d of %d disks connected\n"), + MAX_DISKNO, we.max_disk + 1); } /* Send the close event to OS/400. We DON'T expect a response */ @@ -432,17 +427,15 @@ retry: ((u64)dev_no << 48) | ((u64)flags << 32), 0, 0, 0); if (hvrc != 0) { - printk(VIOD_KERN_WARNING - "bad rc sending event to OS/400 %d\n", (int)hvrc); + pr_warning("bad rc sending event to OS/400 %d\n", (int)hvrc); return 0; } if (d->dev == NULL) { /* this is when we reprobe for new disks */ if (vio_create_viodasd(dev_no) == NULL) { - printk(VIOD_KERN_WARNING - "cannot allocate virtual device for disk %d\n", - dev_no); + pr_warning("cannot allocate virtual device for disk %d\n", + dev_no); return 0; } /* @@ -457,23 +450,20 @@ retry: spin_lock_init(&d->q_lock); q = blk_init_queue(do_viodasd_request, &d->q_lock); if (q == NULL) { - printk(VIOD_KERN_WARNING "cannot allocate queue for disk %d\n", - dev_no); + pr_warning("cannot allocate queue for disk %d\n", dev_no); return 0; } g = alloc_disk(1 << PARTITION_SHIFT); if (g == NULL) { - printk(VIOD_KERN_WARNING - "cannot allocate disk structure for disk %d\n", - dev_no); + pr_warning("cannot allocate disk structure for disk %d\n", + dev_no); blk_cleanup_queue(q); return 0; } d->disk = g; - blk_queue_max_hw_segments(q, VIOMAXBLOCKDMA); - blk_queue_max_phys_segments(q, VIOMAXBLOCKDMA); - blk_queue_max_sectors(q, VIODASD_MAXSECTORS); + blk_queue_max_segments(q, VIOMAXBLOCKDMA); + blk_queue_max_hw_sectors(q, VIODASD_MAXSECTORS); g->major = VIODASD_MAJOR; g->first_minor = dev_no << PARTITION_SHIFT; if (dev_no >= 26) @@ -489,13 +479,12 @@ retry: g->driverfs_dev = d->dev; set_capacity(g, d->size >> 9); - printk(VIOD_KERN_INFO "disk %d: %lu sectors (%lu MB) " - "CHS=%d/%d/%d sector size %d%s\n", - dev_no, (unsigned long)(d->size >> 9), - (unsigned long)(d->size >> 20), - (int)d->cylinders, (int)d->tracks, - (int)d->sectors, (int)d->bytes_per_sector, - d->read_only ? " (RO)" : ""); + pr_info("disk %d: %lu sectors (%lu MB) CHS=%d/%d/%d sector size %d%s\n", + dev_no, (unsigned long)(d->size >> 9), + (unsigned long)(d->size >> 20), + (int)d->cylinders, (int)d->tracks, + (int)d->sectors, (int)d->bytes_per_sector, + d->read_only ? " (RO)" : ""); /* register us in the global list */ add_disk(g); @@ -580,8 +569,8 @@ static int viodasd_handle_read_write(struct vioblocklpevent *bevent) if (error) { const struct vio_error_entry *err; err = vio_lookup_rc(viodasd_err_table, bevent->sub_result); - printk(VIOD_KERN_WARNING "read/write error %d:0x%04x (%s)\n", - event->xRc, bevent->sub_result, err->msg); + pr_warning("read/write error %d:0x%04x (%s)\n", + event->xRc, bevent->sub_result, err->msg); num_sect = blk_rq_sectors(req); } qlock = req->q->queue_lock; @@ -606,8 +595,7 @@ static void handle_block_event(struct HvLpEvent *event) return; /* First, we should NEVER get an int here...only acks */ if (hvlpevent_is_int(event)) { - printk(VIOD_KERN_WARNING - "Yikes! got an int in viodasd event handler!\n"); + pr_warning("Yikes! got an int in viodasd event handler!\n"); if (hvlpevent_need_ack(event)) { event->xRc = HvLpEvent_Rc_InvalidSubtype; HvCallEvent_ackLpEvent(event); @@ -650,7 +638,7 @@ static void handle_block_event(struct HvLpEvent *event) break; default: - printk(VIOD_KERN_WARNING "invalid subtype!"); + pr_warning("invalid subtype!"); if (hvlpevent_need_ack(event)) { event->xRc = HvLpEvent_Rc_InvalidSubtype; HvCallEvent_ackLpEvent(event); @@ -739,29 +727,26 @@ static int __init viodasd_init(void) vio_set_hostlp(); if (viopath_hostLp == HvLpIndexInvalid) { - printk(VIOD_KERN_WARNING "invalid hosting partition\n"); + pr_warning("invalid hosting partition\n"); rc = -EIO; goto early_fail; } - printk(VIOD_KERN_INFO "vers " VIOD_VERS ", hosting partition %d\n", - viopath_hostLp); + pr_info("vers " VIOD_VERS ", hosting partition %d\n", viopath_hostLp); /* register the block device */ rc = register_blkdev(VIODASD_MAJOR, VIOD_GENHD_NAME); if (rc) { - printk(VIOD_KERN_WARNING - "Unable to get major number %d for %s\n", - VIODASD_MAJOR, VIOD_GENHD_NAME); + pr_warning("Unable to get major number %d for %s\n", + VIODASD_MAJOR, VIOD_GENHD_NAME); goto early_fail; } /* Actually open the path to the hosting partition */ rc = viopath_open(viopath_hostLp, viomajorsubtype_blockio, VIOMAXREQ + 2); if (rc) { - printk(VIOD_KERN_WARNING - "error opening path to host partition %d\n", - viopath_hostLp); + pr_warning("error opening path to host partition %d\n", + viopath_hostLp); goto unregister_blk; } @@ -770,7 +755,7 @@ static int __init viodasd_init(void) rc = vio_register_driver(&viodasd_driver); if (rc) { - printk(VIOD_KERN_WARNING "vio_register_driver failed\n"); + pr_warning("vio_register_driver failed\n"); goto unset_handler; } diff --git a/drivers/block/virtio_blk.c b/drivers/block/virtio_blk.c index 51042f0ba7e1..258bc2ae2885 100644 --- a/drivers/block/virtio_blk.c +++ b/drivers/block/virtio_blk.c @@ -1,5 +1,6 @@ //#define DEBUG #include <linux/spinlock.h> +#include <linux/slab.h> #include <linux/blkdev.h> #include <linux/hdreg.h> #include <linux/virtio.h> @@ -49,7 +50,7 @@ static void blk_done(struct virtqueue *vq) unsigned long flags; spin_lock_irqsave(&vblk->lock, flags); - while ((vbr = vblk->vq->vq_ops->get_buf(vblk->vq, &len)) != NULL) { + while ((vbr = virtqueue_get_buf(vblk->vq, &len)) != NULL) { int error; switch (vbr->status) { @@ -69,6 +70,8 @@ static void blk_done(struct virtqueue *vq) vbr->req->sense_len = vbr->in_hdr.sense_len; vbr->req->errors = vbr->in_hdr.errors; } + if (blk_special_request(vbr->req)) + vbr->req->errors = (error != 0); __blk_end_request_all(vbr->req, error); list_del(&vbr->list); @@ -102,6 +105,11 @@ static bool do_req(struct request_queue *q, struct virtio_blk *vblk, vbr->out_hdr.sector = 0; vbr->out_hdr.ioprio = req_get_ioprio(vbr->req); break; + case REQ_TYPE_SPECIAL: + vbr->out_hdr.type = VIRTIO_BLK_T_GET_ID; + vbr->out_hdr.sector = 0; + vbr->out_hdr.ioprio = req_get_ioprio(vbr->req); + break; case REQ_TYPE_LINUX_BLOCK: if (req->cmd[0] == REQ_LB_OP_FLUSH) { vbr->out_hdr.type = VIRTIO_BLK_T_FLUSH; @@ -150,7 +158,7 @@ static bool do_req(struct request_queue *q, struct virtio_blk *vblk, } } - if (vblk->vq->vq_ops->add_buf(vblk->vq, vblk->sg, out, in, vbr) < 0) { + if (virtqueue_add_buf(vblk->vq, vblk->sg, out, in, vbr) < 0) { mempool_free(vbr, vblk->pool); return false; } @@ -179,7 +187,7 @@ static void do_virtblk_request(struct request_queue *q) } if (issued) - vblk->vq->vq_ops->kick(vblk->vq); + virtqueue_kick(vblk->vq); } static void virtblk_prepare_flush(struct request_queue *q, struct request *req) @@ -188,12 +196,45 @@ static void virtblk_prepare_flush(struct request_queue *q, struct request *req) req->cmd[0] = REQ_LB_OP_FLUSH; } +/* return id (s/n) string for *disk to *id_str + */ +static int virtblk_get_id(struct gendisk *disk, char *id_str) +{ + struct virtio_blk *vblk = disk->private_data; + struct request *req; + struct bio *bio; + + bio = bio_map_kern(vblk->disk->queue, id_str, VIRTIO_BLK_ID_BYTES, + GFP_KERNEL); + if (IS_ERR(bio)) + return PTR_ERR(bio); + + req = blk_make_request(vblk->disk->queue, bio, GFP_KERNEL); + if (IS_ERR(req)) { + bio_put(bio); + return PTR_ERR(req); + } + + req->cmd_type = REQ_TYPE_SPECIAL; + return blk_execute_rq(vblk->disk->queue, vblk->disk, req, false); +} + static int virtblk_ioctl(struct block_device *bdev, fmode_t mode, unsigned cmd, unsigned long data) { struct gendisk *disk = bdev->bd_disk; struct virtio_blk *vblk = disk->private_data; + if (cmd == 0x56424944) { /* 'VBID' */ + void __user *usr_data = (void __user *)data; + char id_str[VIRTIO_BLK_ID_BYTES]; + int err; + + err = virtblk_get_id(disk, id_str); + if (!err && copy_to_user(usr_data, id_str, VIRTIO_BLK_ID_BYTES)) + err = -EFAULT; + return err; + } /* * Only allow the generic SCSI ioctls if the host can support it. */ @@ -243,10 +284,12 @@ static int index_to_minor(int index) static int __devinit virtblk_probe(struct virtio_device *vdev) { struct virtio_blk *vblk; + struct request_queue *q; int err; u64 cap; - u32 v; - u32 blk_size, sg_elems; + u32 v, blk_size, sg_elems, opt_io_size; + u16 min_io_size; + u8 physical_block_exp, alignment_offset; if (index_to_minor(index) >= 1 << MINORBITS) return -ENOSPC; @@ -255,7 +298,9 @@ static int __devinit virtblk_probe(struct virtio_device *vdev) err = virtio_config_val(vdev, VIRTIO_BLK_F_SEG_MAX, offsetof(struct virtio_blk_config, seg_max), &sg_elems); - if (err) + + /* We need at least one SG element, whatever they say. */ + if (err || !sg_elems) sg_elems = 1; /* We need an extra sg elements at head and tail. */ @@ -293,13 +338,13 @@ static int __devinit virtblk_probe(struct virtio_device *vdev) goto out_mempool; } - vblk->disk->queue = blk_init_queue(do_virtblk_request, &vblk->lock); - if (!vblk->disk->queue) { + q = vblk->disk->queue = blk_init_queue(do_virtblk_request, &vblk->lock); + if (!q) { err = -ENOMEM; goto out_put_disk; } - vblk->disk->queue->queuedata = vblk; + q->queuedata = vblk; if (index < 26) { sprintf(vblk->disk->disk_name, "vd%c", 'a' + index % 26); @@ -323,10 +368,10 @@ static int __devinit virtblk_probe(struct virtio_device *vdev) /* If barriers are supported, tell block layer that queue is ordered */ if (virtio_has_feature(vdev, VIRTIO_BLK_F_FLUSH)) - blk_queue_ordered(vblk->disk->queue, QUEUE_ORDERED_DRAIN_FLUSH, + blk_queue_ordered(q, QUEUE_ORDERED_DRAIN_FLUSH, virtblk_prepare_flush); else if (virtio_has_feature(vdev, VIRTIO_BLK_F_BARRIER)) - blk_queue_ordered(vblk->disk->queue, QUEUE_ORDERED_TAG, NULL); + blk_queue_ordered(q, QUEUE_ORDERED_TAG, NULL); /* If disk is read-only in the host, the guest should obey */ if (virtio_has_feature(vdev, VIRTIO_BLK_F_RO)) @@ -345,14 +390,13 @@ static int __devinit virtblk_probe(struct virtio_device *vdev) set_capacity(vblk->disk, cap); /* We can handle whatever the host told us to handle. */ - blk_queue_max_phys_segments(vblk->disk->queue, vblk->sg_elems-2); - blk_queue_max_hw_segments(vblk->disk->queue, vblk->sg_elems-2); + blk_queue_max_segments(q, vblk->sg_elems-2); /* No need to bounce any requests */ - blk_queue_bounce_limit(vblk->disk->queue, BLK_BOUNCE_ANY); + blk_queue_bounce_limit(q, BLK_BOUNCE_ANY); /* No real sector limit. */ - blk_queue_max_sectors(vblk->disk->queue, -1U); + blk_queue_max_hw_sectors(q, -1U); /* Host can optionally specify maximum segment size and number of * segments. */ @@ -360,16 +404,45 @@ static int __devinit virtblk_probe(struct virtio_device *vdev) offsetof(struct virtio_blk_config, size_max), &v); if (!err) - blk_queue_max_segment_size(vblk->disk->queue, v); + blk_queue_max_segment_size(q, v); else - blk_queue_max_segment_size(vblk->disk->queue, -1U); + blk_queue_max_segment_size(q, -1U); /* Host can optionally specify the block size of the device */ err = virtio_config_val(vdev, VIRTIO_BLK_F_BLK_SIZE, offsetof(struct virtio_blk_config, blk_size), &blk_size); if (!err) - blk_queue_logical_block_size(vblk->disk->queue, blk_size); + blk_queue_logical_block_size(q, blk_size); + else + blk_size = queue_logical_block_size(q); + + /* Use topology information if available */ + err = virtio_config_val(vdev, VIRTIO_BLK_F_TOPOLOGY, + offsetof(struct virtio_blk_config, physical_block_exp), + &physical_block_exp); + if (!err && physical_block_exp) + blk_queue_physical_block_size(q, + blk_size * (1 << physical_block_exp)); + + err = virtio_config_val(vdev, VIRTIO_BLK_F_TOPOLOGY, + offsetof(struct virtio_blk_config, alignment_offset), + &alignment_offset); + if (!err && alignment_offset) + blk_queue_alignment_offset(q, blk_size * alignment_offset); + + err = virtio_config_val(vdev, VIRTIO_BLK_F_TOPOLOGY, + offsetof(struct virtio_blk_config, min_io_size), + &min_io_size); + if (!err && min_io_size) + blk_queue_io_min(q, blk_size * min_io_size); + + err = virtio_config_val(vdev, VIRTIO_BLK_F_TOPOLOGY, + offsetof(struct virtio_blk_config, opt_io_size), + &opt_io_size); + if (!err && opt_io_size) + blk_queue_io_opt(q, blk_size * opt_io_size); + add_disk(vblk->disk); return 0; @@ -404,7 +477,7 @@ static void __devexit virtblk_remove(struct virtio_device *vdev) kfree(vblk); } -static struct virtio_device_id id_table[] = { +static const struct virtio_device_id id_table[] = { { VIRTIO_ID_BLOCK, VIRTIO_DEV_ANY_ID }, { 0 }, }; @@ -412,7 +485,7 @@ static struct virtio_device_id id_table[] = { static unsigned int features[] = { VIRTIO_BLK_F_BARRIER, VIRTIO_BLK_F_SEG_MAX, VIRTIO_BLK_F_SIZE_MAX, VIRTIO_BLK_F_GEOMETRY, VIRTIO_BLK_F_RO, VIRTIO_BLK_F_BLK_SIZE, - VIRTIO_BLK_F_SCSI, VIRTIO_BLK_F_FLUSH + VIRTIO_BLK_F_SCSI, VIRTIO_BLK_F_FLUSH, VIRTIO_BLK_F_TOPOLOGY }; /* diff --git a/drivers/block/xd.c b/drivers/block/xd.c index d1fd032e7514..18a80ff57ce8 100644 --- a/drivers/block/xd.c +++ b/drivers/block/xd.c @@ -49,6 +49,7 @@ #include <linux/blkpg.h> #include <linux/delay.h> #include <linux/io.h> +#include <linux/gfp.h> #include <asm/system.h> #include <asm/uaccess.h> @@ -242,7 +243,7 @@ static int __init xd_init(void) } /* xd_maxsectors depends on controller - so set after detection */ - blk_queue_max_sectors(xd_queue, xd_maxsectors); + blk_queue_max_hw_sectors(xd_queue, xd_maxsectors); for (i = 0; i < xd_drives; i++) add_disk(xd_gendisk[i]); diff --git a/drivers/block/xen-blkfront.c b/drivers/block/xen-blkfront.c index 05a31e55d278..82ed403147c0 100644 --- a/drivers/block/xen-blkfront.c +++ b/drivers/block/xen-blkfront.c @@ -40,6 +40,7 @@ #include <linux/hdreg.h> #include <linux/cdrom.h> #include <linux/module.h> +#include <linux/slab.h> #include <linux/scatterlist.h> #include <xen/xen.h> @@ -346,15 +347,14 @@ static int xlvbd_init_blk_queue(struct gendisk *gd, u16 sector_size) /* Hard sector size and max sectors impersonate the equiv. hardware. */ blk_queue_logical_block_size(rq, sector_size); - blk_queue_max_sectors(rq, 512); + blk_queue_max_hw_sectors(rq, 512); /* Each segment in a request is up to an aligned page in size. */ blk_queue_segment_boundary(rq, PAGE_SIZE - 1); blk_queue_max_segment_size(rq, PAGE_SIZE); /* Ensure a merged request will fit in a single I/O ring slot. */ - blk_queue_max_phys_segments(rq, BLKIF_MAX_SEGMENTS_PER_REQUEST); - blk_queue_max_hw_segments(rq, BLKIF_MAX_SEGMENTS_PER_REQUEST); + blk_queue_max_segments(rq, BLKIF_MAX_SEGMENTS_PER_REQUEST); /* Make sure buffer addresses are sector-aligned. */ blk_queue_dma_alignment(rq, 511); @@ -1050,7 +1050,7 @@ static const struct block_device_operations xlvbd_block_fops = }; -static struct xenbus_device_id blkfront_ids[] = { +static const struct xenbus_device_id blkfront_ids[] = { { "vbd" }, { "" } }; diff --git a/drivers/block/xsysace.c b/drivers/block/xsysace.c index e5c5415eb45e..a7b83c0a7eb5 100644 --- a/drivers/block/xsysace.c +++ b/drivers/block/xsysace.c @@ -1198,10 +1198,10 @@ ace_of_probe(struct of_device *op, const struct of_device_id *match) dev_dbg(&op->dev, "ace_of_probe(%p, %p)\n", op, match); /* device id */ - id = of_get_property(op->node, "port-number", NULL); + id = of_get_property(op->dev.of_node, "port-number", NULL); /* physaddr */ - rc = of_address_to_resource(op->node, 0, &res); + rc = of_address_to_resource(op->dev.of_node, 0, &res); if (rc) { dev_err(&op->dev, "invalid address\n"); return rc; @@ -1209,11 +1209,11 @@ ace_of_probe(struct of_device *op, const struct of_device_id *match) physaddr = res.start; /* irq */ - irq = irq_of_parse_and_map(op->node, 0); + irq = irq_of_parse_and_map(op->dev.of_node, 0); /* bus width */ bus_width = ACE_BUS_WIDTH_16; - if (of_find_property(op->node, "8-bit", NULL)) + if (of_find_property(op->dev.of_node, "8-bit", NULL)) bus_width = ACE_BUS_WIDTH_8; /* Call the bus-independant setup code */ @@ -1227,7 +1227,7 @@ static int __devexit ace_of_remove(struct of_device *op) } /* Match table for of_platform binding */ -static struct of_device_id ace_of_match[] __devinitdata = { +static const struct of_device_id ace_of_match[] __devinitconst = { { .compatible = "xlnx,opb-sysace-1.00.b", }, { .compatible = "xlnx,opb-sysace-1.00.c", }, { .compatible = "xlnx,xps-sysace-1.00.a", }, @@ -1237,13 +1237,12 @@ static struct of_device_id ace_of_match[] __devinitdata = { MODULE_DEVICE_TABLE(of, ace_of_match); static struct of_platform_driver ace_of_driver = { - .owner = THIS_MODULE, - .name = "xsysace", - .match_table = ace_of_match, .probe = ace_of_probe, .remove = __devexit_p(ace_of_remove), .driver = { .name = "xsysace", + .owner = THIS_MODULE, + .of_match_table = ace_of_match, }, }; diff --git a/drivers/block/z2ram.c b/drivers/block/z2ram.c index 64f941e0f14b..9114654b54d9 100644 --- a/drivers/block/z2ram.c +++ b/drivers/block/z2ram.c @@ -33,6 +33,7 @@ #include <linux/module.h> #include <linux/blkdev.h> #include <linux/bitops.h> +#include <linux/slab.h> #include <asm/setup.h> #include <asm/amigahw.h> |