summaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAgeFilesLines
* Merge branch 'for-linus' of git://neil.brown.name/mdLinus Torvalds2009-01-0813-247/+435
|\ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | * 'for-linus' of git://neil.brown.name/md: md: don't retry recovery of raid1 that fails due to error on source drive. md: Allow md devices to be created by name. md: make devices disappear when they are no longer needed. md: centralise all freeing of an 'mddev' in 'md_free' md: move allocation of ->queue from mddev_find to md_probe md: need another print_sb for mdp_superblock_1 md: use list_for_each_entry macro directly md: raid0: make hash_spacing and preshift sector-based. md: raid0: Represent the size of strip zones in sectors. md: raid0 create_strip_zones(): Add KERN_INFO/KERN_ERR to printk's. md: raid0 create_strip_zones(): Make two local variables sector-based. md: raid0: Represent zone->zone_offset in sectors. md: raid0: Represent device offset in sectors. md: raid0_make_request(): Replace local variable block by sector. md: raid0_make_request(): Remove local variable chunk_size. md: raid0_make_request(): Replace chunksize_bits by chunksect_bits. md: use sysfs_notify_dirent to notify changes to md/sync_action. md: fix bitmap-on-external-file bug.
| * md: don't retry recovery of raid1 that fails due to error on source drive.NeilBrown2009-01-093-3/+13
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | If a raid1 has only one working drive and it has a sector which gives an error on read, then an attempt to recover onto a spare will fail, but as the single remaining drive is not removed from the array, the recovery will be immediately re-attempted, resulting in an infinite recovery loop. So detect this situation and don't retry recovery once an error on the lone remaining drive is detected. Allow recovery to be retried once every time a spare is added in case the problem wasn't actually a media error. Signed-off-by: NeilBrown <neilb@suse.de>
| * md: Allow md devices to be created by name.NeilBrown2009-01-092-18/+102
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Using sequential numbers to identify md devices is somewhat artificial. Using names can be a lot more user-friendly. Also, creating md devices by opening the device special file is a bit awkward. So this patch provides a new option for creating and naming devices. Writing a name such as "md_home" to /sys/modules/md_mod/parameters/new_array will cause an array with that name to be created. It will appear in /sys/block/ /proc/partitions and /proc/mdstat as 'md_home'. It will have an arbitrary minor number allocated. md devices that a created by an open are destroyed on the last close when the device is inactive. For named md devices, they will not be destroyed until the array is explicitly stopped, either with the STOP_ARRAY ioctl or by writing 'clear' to /sys/block/md_XXXX/md/array_state. The name of the array must start 'md_' to avoid conflict with other devices. Signed-off-by: NeilBrown <neilb@suse.de>
| * md: make devices disappear when they are no longer needed.NeilBrown2009-01-093-12/+67
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Currently md devices, once created, never disappear until the module is unloaded. This is essentially because the gendisk holds a reference to the mddev, and the mddev holds a reference to the gendisk, this a circular reference. If we drop the reference from mddev to gendisk, then we need to ensure that the mddev is destroyed when the gendisk is destroyed. However it is not possible to hook into the gendisk destruction process to enable this. So we drop the reference from the gendisk to the mddev and destroy the gendisk when the mddev gets destroyed. However this has a complication. Between the call __blkdev_get->get_gendisk->kobj_lookup->md_probe and the call __blkdev_get->md_open there is no obvious way to hold a reference on the mddev any more, so unless something is done, it will disappear and gendisk will be destroyed prematurely. Also, once we decide to destroy the mddev, there will be an unlockable moment before the gendisk is unlinked (blk_unregister_region) during which a new reference to the gendisk can be created. We need to ensure that this reference can not be used. i.e. the ->open must fail. So: 1/ in md_probe we set a flag in the mddev (hold_active) which indicates that the array should be treated as active, even though there are no references, and no appearance of activity. This is cleared by md_release when the device is closed if it is no longer needed. This ensures that the gendisk will survive between md_probe and md_open. 2/ In md_open we check if the mddev we expect to open matches the gendisk that we did open. If there is a mismatch we return -ERESTARTSYS and modify __blkdev_get to retry from the top in that case. In the -ERESTARTSYS sys case we make sure to wait until the old gendisk (that we succeeded in opening) is really gone so we loop at most once. Some udev configurations will always open an md device when it first appears. If we allow an md device that was just created by an open to disappear on an immediate close, then this can race with such udev configurations and result in an infinite loop the device being opened and closed, then re-open due to the 'ADD' even from the first open, and then close and so on. So we make sure an md device, once created by an open, remains active at least until some md 'ioctl' has been made on it. This means that all normal usage of md devices will allow them to disappear promptly when not needed, but the worst that an incorrect usage will do it cause an inactive md device to be left in existence (it can easily be removed). As an array can be stopped by writing to a sysfs attribute echo clear > /sys/block/mdXXX/md/array_state we need to use scheduled work for deleting the gendisk and other kobjects. This allows us to wait for any pending gendisk deletion to complete by simply calling flush_scheduled_work(). Signed-off-by: NeilBrown <neilb@suse.de>
| * md: centralise all freeing of an 'mddev' in 'md_free'NeilBrown2009-01-091-9/+11
| | | | | | | | | | | | | | | | | | md_free is the .release handler for the md kobj_type. So it makes sense to release all the objects referenced by the mddev in there, rather than just prior to calling kobject_put for what we think is the last time. Signed-off-by: NeilBrown <neilb@suse.de>
| * md: move allocation of ->queue from mddev_find to md_probeNeilBrown2009-01-091-11/+17
| | | | | | | | | | | | | | | | | | | | It is more balanced to just do simple initialisation in mddev_find, which allocates and links a new md device, and leave all the more sophisticated allocation to md_probe (which calls mddev_find). md_probe already allocated the gendisk. It should allocate the queue too. Signed-off-by: NeilBrown <neilb@suse.de>
| * md: need another print_sb for mdp_superblock_1Cheng Renquan2009-01-092-6/+62
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | md_print_devices is called in two code path: MD_BUG(...), and md_ioctl with PRINT_RAID_DEBUG. it will dump out all in use md devices information; However, it wrongly processed two types of superblock in one: The header file <linux/raid/md_p.h> has defined two types of superblock, struct mdp_superblock_s (typedefed with mdp_super_t) according to md with metadata 0.90, and struct mdp_superblock_1 according to md with metadata 1.0 and later, These two types of superblock are very different, The md_print_devices code processed them both in mdp_super_t, that would lead to wrong informaton dump like: [ 6742.345877] [ 6742.345887] md: ********************************** [ 6742.345890] md: * <COMPLETE RAID STATE PRINTOUT> * [ 6742.345892] md: ********************************** [ 6742.345896] md1: <ram7><ram6><ram5><ram4> [ 6742.345907] md: rdev ram7, SZ:00065472 F:0 S:1 DN:3 [ 6742.345909] md: rdev superblock: [ 6742.345914] md: SB: (V:0.90.0) ID:<42ef13c7.598c059a.5f9f1645.801e9ee6> CT:4919856d [ 6742.345918] md: L5 S00065472 ND:4 RD:4 md1 LO:2 CS:65536 [ 6742.345922] md: UT:4919856d ST:1 AD:4 WD:4 FD:0 SD:0 CSUM:b7992907 E:00000001 [ 6742.345924] D 0: DISK<N:0,(1,8),R:0,S:6> [ 6742.345930] D 1: DISK<N:1,(1,10),R:1,S:6> [ 6742.345933] D 2: DISK<N:2,(1,12),R:2,S:6> [ 6742.345937] D 3: DISK<N:3,(1,14),R:3,S:6> [ 6742.345942] md: THIS: DISK<N:3,(1,14),R:3,S:6> ... [ 6742.346058] md0: <ram3><ram2><ram1><ram0> [ 6742.346067] md: rdev ram3, SZ:00065472 F:0 S:1 DN:3 [ 6742.346070] md: rdev superblock: [ 6742.346073] md: SB: (V:1.0.0) ID:<369aad81.00000000.00000000.00000000> CT:9a322a9c [ 6742.346077] md: L-1507699579 S976570180 ND:48 RD:0 md0 LO:65536 CS:196610 [ 6742.346081] md: UT:00000018 ST:0 AD:131048 WD:0 FD:8 SD:0 CSUM:00000000 E:00000000 [ 6742.346084] D 0: DISK<N:-1,(-1,-1),R:-1,S:-1> [ 6742.346089] D 1: DISK<N:-1,(-1,-1),R:-1,S:-1> [ 6742.346092] D 2: DISK<N:-1,(-1,-1),R:-1,S:-1> [ 6742.346096] D 3: DISK<N:-1,(-1,-1),R:-1,S:-1> [ 6742.346102] md: THIS: DISK<N:0,(0,0),R:0,S:0> ... [ 6742.346219] md: ********************************** [ 6742.346221] Here md1 is metadata 0.90.0, and md0 is metadata 1.2 After some more code to distinguish these two types of superblock, in this patch, it will generate dump information like: [ 7906.755790] [ 7906.755799] md: ********************************** [ 7906.755802] md: * <COMPLETE RAID STATE PRINTOUT> * [ 7906.755804] md: ********************************** [ 7906.755808] md1: <ram7><ram6><ram5><ram4> [ 7906.755819] md: rdev ram7, SZ:00065472 F:0 S:1 DN:3 [ 7906.755821] md: rdev superblock (MJ:0): [ 7906.755826] md: SB: (V:0.90.0) ID:<3fca7a0d.a612bfed.5f9f1645.801e9ee6> CT:491989f3 [ 7906.755830] md: L5 S00065472 ND:4 RD:4 md1 LO:2 CS:65536 [ 7906.755834] md: UT:491989f3 ST:1 AD:4 WD:4 FD:0 SD:0 CSUM:00fb52ad E:00000001 [ 7906.755836] D 0: DISK<N:0,(1,8),R:0,S:6> [ 7906.755842] D 1: DISK<N:1,(1,10),R:1,S:6> [ 7906.755845] D 2: DISK<N:2,(1,12),R:2,S:6> [ 7906.755849] D 3: DISK<N:3,(1,14),R:3,S:6> [ 7906.755855] md: THIS: DISK<N:3,(1,14),R:3,S:6> ... [ 7906.755972] md0: <ram3><ram2><ram1><ram0> [ 7906.755981] md: rdev ram3, SZ:00065472 F:0 S:1 DN:3 [ 7906.755984] md: rdev superblock (MJ:1): [ 7906.755989] md: SB: (V:1) (F:0) Array-ID:<5fbcf158:55aa:5fbe:9a79:1e939880dcbd> [ 7906.755990] md: Name: "DG5:0" CT:1226410480 [ 7906.755998] md: L5 SZ130944 RD:4 LO:2 CS:128 DO:24 DS:131048 SO:8 RO:0 [ 7906.755999] md: Dev:00000003 UUID: 9194d744:87f7:a448:85f2:7497b84ce30a [ 7906.756001] md: (F:0) UT:1226410480 Events:0 ResyncOffset:-1 CSUM:0dbcd829 [ 7906.756003] md: (MaxDev:384) ... [ 7906.756113] md: ********************************** [ 7906.756116] this md0 (metadata 1.2) information dumping is exactly according to struct mdp_superblock_1. Signed-off-by: Cheng Renquan <crquan@gmail.com> Cc: Neil Brown <neilb@suse.de> Cc: Dan Williams <dan.j.williams@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: NeilBrown <neilb@suse.de>
| * md: use list_for_each_entry macro directlyCheng Renquan2009-01-0910-94/+60
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The rdev_for_each macro defined in <linux/raid/md_k.h> is identical to list_for_each_entry_safe, from <linux/list.h>, it should be defined to use list_for_each_entry_safe, instead of reinventing the wheel. But some calls to each_entry_safe don't really need a safe version, just a direct list_for_each_entry is enough, this could save a temp variable (tmp) in every function that used rdev_for_each. In this patch, most rdev_for_each loops are replaced by list_for_each_entry, totally save many tmp vars; and only in the other situations that will call list_del to delete an entry, the safe version is used. Signed-off-by: Cheng Renquan <crquan@gmail.com> Signed-off-by: NeilBrown <neilb@suse.de>
| * md: raid0: make hash_spacing and preshift sector-based.Andre Noll2009-01-092-31/+31
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This patch renames the hash_spacing and preshift members of struct raid0_private_data to spacing and sector_shift respectively and changes the semantics as follows: We always have spacing = 2 * hash_spacing. In case sizeof(sector_t) > sizeof(u32) we also have sector_shift = preshift + 1 while sector_shift = preshift = 0 otherwise. Note that the values of nb_zone and zone are unaffected by these changes because in the sector_div() preceeding the assignement of these two variables both arguments double. Signed-off-by: Andre Noll <maan@systemlinux.org> Signed-off-by: NeilBrown <neilb@suse.de>
| * md: raid0: Represent the size of strip zones in sectors.Andre Noll2009-01-092-14/+14
| | | | | | | | | | | | | | This completes the block -> sector conversion of struct strip_zone. Signed-off-by: Andre Noll <maan@systemlinux.org> Signed-off-by: NeilBrown <neilb@suse.de>
| * md: raid0 create_strip_zones(): Add KERN_INFO/KERN_ERR to printk's.Andre Noll2009-01-091-21/+25
| | | | | | | | | | | | | | This patch consists only of these trivial changes. Signed-off-by: Andre Noll <maan@systemlinux.org> Signed-off-by: NeilBrown <neilb@suse.de>
| * md: raid0 create_strip_zones(): Make two local variables sector-based.Andre Noll2009-01-091-14/+13
| | | | | | | | | | | | | | | | | | | | | | | | | | current_offset and curr_zone_offset stored the corresponding offsets as 1K quantities. Rename them to current_start and curr_zone_start to match the naming of struct strip_zone and store the offsets as sector counts. Also, add KERN_INFO to the printk() affected by this change to make checkpatch happy. Signed-off-by: Andre Noll <maan@systemlinux.org> Signed-off-by: NeilBrown <neilb@suse.de>
| * md: raid0: Represent zone->zone_offset in sectors.Andre Noll2009-01-092-7/+7
| | | | | | | | | | | | | | | | For the same reason as in the previous patch, rename it from zone_offset to zone_start. Signed-off-by: Andre Noll <maan@systemlinux.org> Signed-off-by: NeilBrown <neilb@suse.de>
| * md: raid0: Represent device offset in sectors.Andre Noll2009-01-092-6/+5
| | | | | | | | | | | | | | | | Rename zone->dev_offset to zone->dev_start to make sure all users have been converted. Signed-off-by: Andre Noll <maan@systemlinux.org> Signed-off-by: NeilBrown <neilb@suse.de>
| * md: raid0_make_request(): Replace local variable block by sector.Andre Noll2009-01-091-7/+6
| | | | | | | | | | | | | | This change already simplifies the code a bit. Signed-off-by: Andre Noll <maan@systemlinux.org> Signed-off-by: NeilBrown <neilb@suse.de>
| * md: raid0_make_request(): Remove local variable chunk_size.Andre Noll2009-01-091-4/+3
| | | | | | | | | | | | | | We might as well use chunk_sects instead. Signed-off-by: Andre Noll <maan@systemlinux.org> Signed-off-by: NeilBrown <neilb@suse.de>
| * md: raid0_make_request(): Replace chunksize_bits by chunksect_bits.Andre Noll2009-01-091-5/+5
| | | | | | | | | | | | | | | | | | | | | | As ffz(~(2 * x)) = ffz(~x) + 1, we have chunksect_bits = chunksize_bits + 1. Fixup all users accordingly. Signed-off-by: Andre Noll <maan@systemlinux.org> Signed-off-by: NeilBrown <neilb@suse.de>
| * md: use sysfs_notify_dirent to notify changes to md/sync_action.NeilBrown2009-01-092-7/+14
| | | | | | | | | | | | | | There is no compelling need for this, but sysfs_notify_dirent is a nicer interface and the change is good for consistency. Signed-off-by: NeilBrown <neilb@suse.de>
| * md: fix bitmap-on-external-file bug.NeilBrown2009-01-091-3/+5
| | | | | | | | | | | | | | | | | | | | | | commit a2ed9615e3222645007fc19991aedf30eed3ecfd fixed a bug with 'internal' bitmaps, but in the process broke 'in a file' bitmaps. So they are broken in 2.6.28 This fixes it, and needs to go in 2.6.28-stable. Signed-off-by: NeilBrown <neilb@suse.de> Cc: stable@kernel.org
* | Merge branch 'upstream-linus' of ↵Linus Torvalds2009-01-0813-203/+336
|\ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | git://git.kernel.org/pub/scm/linux/kernel/git/jgarzik/libata-dev * 'upstream-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jgarzik/libata-dev: libata: clean up the SFF code for coding style libata: Add 32bit PIO support [libata] ahci: Withdraw IGN_SERR_INTERNAL for SB800 SATA pata_hpt366: reimplement mode programming [libata] pata_hpt3x3: correct _freeze() function declaration libata: Add special ata_pio_need_iordy() handling for Compact Flash. pata_platform: __pata_platform_remove() shouldn't be in discard section sata_sil24: remove unused sil24_port_multiplier [libata] ahci: Add SATA GEN3 related messages ata_piix: save, use saved and restore IOCFG pata_ali: Fix and workaround for FIFO DMA bug pata_ali: force initialise a few bits pata_hpt3x3: Workarounds for chipset
| * | libata: clean up the SFF code for coding styleAlan Cox2009-01-081-65/+67
| | | | | | | | | | | | | | | Signed-off-by: Alan Cox <alan@redhat.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
| * | libata: Add 32bit PIO supportAlan Cox2009-01-087-9/+66
| | | | | | | | | | | | | | | | | | | | | | | | | | | This matters for some controllers and in one or two cases almost doubles PIO performance. Add a bmdma32 operations set we can inherit and activate it for some controllers Signed-off-by: Alan Cox <alan@redhat.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
| * | [libata] ahci: Withdraw IGN_SERR_INTERNAL for SB800 SATAShane Huang2009-01-081-2/+6
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | There is an issue in ATI SB600/SB700 SATA that PxSERR.E should not be set on some conditions, which will lead to many SATA ODD error messages. commit 55a61604cd1354e1783364e1c901034f2f474b7d is the workaround. Since SB800 fixed this HW issue, IGN_SERR_INTERNAL should be withdrawn for SB800. Signed-off-by: Shane Huang <shane.huang@amd.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
| * | pata_hpt366: reimplement mode programmingTejun Heo2009-01-081-65/+44
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Reimplement mode programming logic of pata_hpt366 such that it's identical to that of IDE hpt366 driver. The differences were... * pata_hpt366 used 0xCFFF8FFFF to mask pio modes and 0x3FFFFFFF dma modes. IDE hpt366 uses 0xC1F8FFFF for PIO, 0x303800FF for MWDMA and 0x30070000 for UDMA. * pata_hpt366 doesn't set 0x08000000 for PIO unless it's already set and always turns it on for MWDMA/UDMA. IDE hpt366 doesn't bother with the bit. It always uses what was there. * IDE hpt366 always clears 0xC0000000. pata_hpt366 doesn't. Signed-off-by: Tejun Heo <tj@kernel.org> Cc: Alan Cox <alan@lxorguk.ukuu.org.uk> Cc: Sergei Shtylyov <sshtylyov@ru.mvista.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
| * | [libata] pata_hpt3x3: correct _freeze() function declarationJeff Garzik2009-01-081-2/+2
| | | | | | | | | | | | Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
| * | libata: Add special ata_pio_need_iordy() handling for Compact Flash.David Daney2009-01-081-0/+4
| | | | | | | | | | | | | | | | | | | | | | | | According to the Compact Flash specification r4.1, PIO modes 5 and 6 do not use iordy. Signed-off-by: David Daney <ddaney@caviumnetworks.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
| * | pata_platform: __pata_platform_remove() shouldn't be in discard sectionSonic Zhang2009-01-081-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | -- UPD include/linux/compile.h `___pata_platform_remove' referenced in section `__ksymtab_gpl' of drivers/built-in.o: defined in discarded section `.devexit.text' of drivers/built-in.o make: *** [.tmp_vmlinux1] Error 1 -- __pata_platform_remove() should not be in discarded section __pata_platform_remove(struct device *dev) is invoked in both pata_platform.c and pata_of_platform.c by reomve function defined in discarded section ".devexit.text". An exported function should not be put into discarded section. Signed-off-by: Sonic Zhang <sonic.zhang@analog.com> Signed-off-by: Bryan Wu <cooloney@kernel.org> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
| * | sata_sil24: remove unused sil24_port_multiplierGrant Grundler2009-01-081-7/+0
| | | | | | | | | | | | | | | | | | | | | | | | AFAICT, struct sil24_port_multiplier isn't used anywhere. Remove it. Signed-off-by: Grant Grundler <grundler@google.com> Signed-off-by: Tejun Heo <tj@kernel.org> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
| * | [libata] ahci: Add SATA GEN3 related messagesShane Huang2009-01-082-0/+3
| | | | | | | | | | | | | | | | | | | | | | | | The present AHCI driver seems to support SATA GEN 3 speed, but the related messages should be modified. Signed-off-by: Shane Huang <shane.huang@amd.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
| * | ata_piix: save, use saved and restore IOCFGTejun Heo2009-01-081-15/+34
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Certain ACPI implementations mess up IOCFG on _STM making libata detect cable type incorrectly after a suspend/resume cycle. This patch makes ata_piix save IOCFG on attach, use the saved value for things which aren't dynamic and restore it on detach so that the next driver also gets the BIOS initialized value. This patch contains the following changes. * makes ich_pata_cable_detect() use saved_iocfg. * make piix_iocfg_bit18_quirk() take @host and use saved_iocfg. * hpriv allocation moved upwards to save iocfg before doing anything else. This fixes bz#11879. Andreas Mohr reported and diagnosed the problem. Signed-off-by: Tejun Heo <tj@kernel.org> Cc: Andreas Mohr <andi@lisas.de> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
| * | pata_ali: Fix and workaround for FIFO DMA bugAlan Cox2009-01-081-2/+30
| | | | | | | | | | | | | | | | | | | | | | | | In very obscure cases this can cause problems. We need to help the hardware out a bit to avoid DMA problems on a reset. Signed-off-by: Alan Cox <alan@redhat.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
| * | pata_ali: force initialise a few bitsAlan Cox2009-01-081-33/+38
| | | | | | | | | | | | | | | | | | | | | We can't assume some of the setup here on non x86 boxes. Signed-off-by: Alan Cox <alan@redhat.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
| * | pata_hpt3x3: Workarounds for chipsetAlan Cox2009-01-081-6/+45
| | | | | | | | | | | | | | | | | | | | | | | | | | | Correct the DMA bit flags (UDMA and MWDMA were swapped) Add workarounds so that we clear ERR and INTR bits before issuing a DMA Add workarounds so that we stop a live DMA before touching the CTL register Signed-off-by: Alan Cox <alan@redhat.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
* | | async: make async_synchronize_full() more serializingArjan van de Ven2009-01-081-1/+3
|/ / | | | | | | | | | | | | | | | | | | | | | | | | turns out that there are real problems with allowing async tasks that are scheduled from async tasks to run after the async_synchronize_full() returns. This patch makes the _full more strict and a complete synchronization. Later I might need to add back a lighter form of synchronization for other uses.. but not right now. Signed-off-by: Arjan van de Ven <arjan@linux.intel.com> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* | Merge branch 'next' of ↵Linus Torvalds2009-01-0874-502/+3089
|\ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | git://git.kernel.org/pub/scm/linux/kernel/git/benh/powerpc * 'next' of git://git.kernel.org/pub/scm/linux/kernel/git/benh/powerpc: (53 commits) serial: Add driver for the Cell Network Processor serial port NWP device powerpc: enable dynamic ftrace powerpc/cell: Fix the prototype of create_vma_map() powerpc/mm: Make clear_fixmap() actually work powerpc/kdump: Use ppc_save_regs() in crash_setup_regs() powerpc: Export cacheable_memzero as its now used in a driver powerpc: Fix missing semicolons in mmu_decl.h powerpc/pasemi: local_irq_save uses an unsigned long powerpc/cell: Fix some u64 vs. long types powerpc/cell: Use correct types in beat files powerpc: Use correct type in prom_init.c powerpc: Remove unnecessary casts mtd/ps3vram: Use _PAGE_NO_CACHE in memory ioremap mtd/ps3vram: Use msleep in waits mtd/ps3vram: Use proper kernel types mtd/ps3vram: Cleanup ps3vram driver messages mtd/ps3vram: Remove ps3vram debug routines mtd/ps3vram: Add modalias support to the ps3vram driver mtd/ps3vram: Add ps3vram driver for accessing video RAM as MTD powerpc: Fix iseries drivers build failure without CONFIG_VIOPATH ...
| * | serial: Add driver for the Cell Network Processor serial port NWP deviceBenjamin Krill2009-01-086-1/+534
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Add support for the nwp serial device which is connected to a DCR bus. It uses the of_serial device driver to determine necessary properties from the device tree. The supported device is added as serial port number 85. NWP stands for network processor and it is part of the QPACE - Quantum Chromodynamics Parallel Computing on the Cell Broadband Engine project. The implementation is a lightweight uart implementation with the focus to consume as little resources as possible and it is connected to a DCR bus. Signed-off-by: Benjamin Krill <ben@codiert.org> Signed-off-by: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
| * | powerpc: enable dynamic ftraceSteven Rostedt2009-01-081-0/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | | This patch enables dynamic ftrace. The PowerPC port was dependent on other code not yet in mainline. Now that the code is, we can now let PowerPC compile with dynamic ftrace. Signed-off-by: Steven Rostedt <srostedt@redhat.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
| * | powerpc/cell: Fix the prototype of create_vma_map()Stephen Rothwell2009-01-081-1/+1
| | | | | | | | | | | | | | | | | | Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au> Acked-by: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
| * | powerpc/mm: Make clear_fixmap() actually workAnton Vorontsov2009-01-081-1/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The clear_fixmap() routine issues map_page() with flags set to 0. Currently this causes a BUG_ON() inside the map_page(), as it assumes that a PTE should be clear before mapping. This patch makes the map_page() to trigger the BUG_ON() only if the flags were set. Signed-off-by: Anton Vorontsov <avorontsov@ru.mvista.com> Acked-by: Kumar Gala <galak@kernel.crashing.org> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
| * | powerpc/kdump: Use ppc_save_regs() in crash_setup_regs()Anton Vorontsov2009-01-081-55/+0
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The patch replaces internal registers dump implementation with ppc_save_regs(). From now on PPC64 and PPC32 are using the same code for crash_setup_regs(). NOTE: The old regs dump implementation was capturing SP (r1) directly as is, so you could see crash_kexec() function on top of the back-trace. But ppc_save_regs() goes up one stack frame, so you'll not see it anymore, at the top-level you'll see who actually triggered the crash dump instead. Signed-off-by: Anton Vorontsov <avorontsov@ru.mvista.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
| * | powerpc: Export cacheable_memzero as its now used in a driverKumar Gala2009-01-081-0/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | The Freescale PowerPC specific gianfar driver (gig-e) uses cacheable_memzero for performance reasons we need to export the symbol to allow the driver to be built as a module. Signed-off-by: Kumar Gala <galak@kernel.crashing.org> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
| * | powerpc: Fix missing semicolons in mmu_decl.hBenjamin Herrenschmidt2009-01-081-3/+3
| | | | | | | | | | | | | | | | | | | | | | | | | | | This is a brown paper bag from one of my earlier patches that breaks build on 40x and 8xx. And yes, I've now added 40x and 8xx to my list of test configs :-) Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
| * | powerpc/pasemi: local_irq_save uses an unsigned longIngo Molnar2009-01-081-1/+1
| | | | | | | | | | | | | | | | | | | | | [Split from a larger patch - sfr] Signed-off-by: Ingo Molnar <mingo@elte.hu> Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
| * | powerpc/cell: Fix some u64 vs. long typesIngo Molnar2009-01-083-5/+5
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | in/out_be64() work on u64s. The first parameter to ppc_md.ioremap is a phys_addr_t. Signed-off-by: Ingo Molnar <mingo@elte.hu> Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au> Acked-by: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
| * | powerpc/cell: Use correct types in beat filesIngo Molnar2009-01-083-14/+15
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Only pass the address of a u64 if that is what the function requires. [Split out of a larger patch - sfr] [update comment - sfr] Signed-off-by: Ingo Molnar <mingo@elte.hu> Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
| * | powerpc: Use correct type in prom_init.cIngo Molnar2009-01-081-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | tce_entryp is a "u64 *" not an "unsigned long *". [Split from a large patch -sfr] Signed-off-by: Ingo Molnar <mingo@elte.hu> Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
| * | powerpc: Remove unnecessary castsStephen Rothwell2009-01-081-7/+7
| | | | | | | | | | | | | | | | | | | | | of_get_flat_dt_prop() returns a "void *", so we don't need to cast when assigning its result to a pointer variable. Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
| * | mtd/ps3vram: Use _PAGE_NO_CACHE in memory ioremapGeoff Levand2009-01-081-6/+8
| | | | | | | | | | | | | | | | | | | | | | | | | | | Use _PAGE_NO_CACHE for gpu memory ioremap. Also, add __iomem attribute to gpu memory pointer and change use of memset() to memset_io(). Signed-off-by: Geoff Levand <geoffrey.levand@am.sony.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
| * | mtd/ps3vram: Use msleep in waitsGeoff Levand2009-01-081-23/+17
| | | | | | | | | | | | | | | | | | | | | | | | Replace the use of udelay() with msleep() in the looping wait routines ps3vram_notifier_wait() and ps3vram_wait_ring(). Signed-off-by: Geoff Levand <geoffrey.levand@am.sony.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
| * | mtd/ps3vram: Use proper kernel typesGeoff Levand2009-01-081-26/+30
| | | | | | | | | | | | | | | | | | | | | | | | Replace the use of stdint.h types with kernel types in the ps3vram driver. Signed-off-by: Geoff Levand <geoffrey.levand@am.sony.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
View Lorry Controller Status