10 Sep, 2005
4 commits
-
The SAS transport class contains common code to deal with SAS HBAs, an
aproximated representation of SAS topologies in the driver model,
and various sysfs attributes to expose these topologies and managment
interfaces to userspace.In addition to the basic SCSI core objects this transport class introduces
two additional intermediate objects: The SAS PHY as represented by struct
sas_phy defines an "outgoing" PHY on a SAS HBA or Expander, and the SAS
remote PHY represented by struct sas_rphy defines an "incoming" PHY on a
SAS Expander or end device. Note that this is purely a software concept, the
underlying hardware for a PHY and a remote PHY is the exactly the same.There is no concept of a SAS port in this code, users can see what PHYs
form a wide port based on the port_identifier attribute, which is the same
for all PHYs in a port.This submission doesn't handle hot-plug addition or removal of SAS devices
and thus doesn't do scanning in a workqueue yet, that will be added in
phase2 after this submission. In a third phase I will add additional
managment infrastructure.I think this submission is ready for 2.6.14, but additional comments are
of course very welcome.I'd like to thanks James Smart a lot for his very useful input on the
design.Signed-off-by: Christoph Hellwig
Signed-off-by: James Bottomley -
The soon to be released smartmontools 5.34 uses the
READ DEFECT DATA command on SCSI disks. A disk that
has defect list entries (or worse, an increasing number
of them) is at risk.Currently the first invocation of smartctl causes this:
scsi: unknown opcode 0x37
message to appear the console and in the log.The READ DEFECT DATA SCSI command does not change
the state of a disk. Its opcode (0x37) is valid for
SBC devices (e.g. disks) and SMC-2 devices (media
changers) where it is called INITIALIZE STATUS ELEMENT
WITH RANGE and again doesn't change the external state
of the device.Changelog:
- mark SCSI opcode 0x37 (READ DEFECT DATA) as
safe_for_readSigned-off-by: Douglas Gilbert
Acked-by: Jens Axboe
Signed-off-by: James Bottomley -
Further to the problem discussed in this post:
http://marc.theaimsgroup.com/?l=linux-scsi&m=112540053711489&w=2It seems that the sg driver does not need to set the VM_IO flag
on pages that it memory maps to the user space since they are
not from the IO space. Ahmed Teirelbar
wants the facility and has tested this patch as I have without
adverse effects.The oops protection is still important. Some users really did
try and use dio transfers from the sg driver to memory mapped
IO space (on a video capture card if my memory serves) during the
lk 2.4 series. I'm not sure how successful it was but that will
now be politely refused in lk 2.6.13+ .Changelog:
- set the page flags for sg's reserved buffer mmap-ed
to the user space to VM_RESERVED (rather than
VM_RESERVED | VM_IO )Signed-off-by: Douglas Gilbert
Signed-off-by: James Bottomley -
Actually, just one problem and one cosmetic fix:
1) We need to dequeue for the loop and kill case (it seems easiest
simply to dequeue in the scsi_kill_request() routine)
2) There's no real need to drop the queue lock. __scsi_done() is lock
agnostic, so since there's no requirement, let's just leave it in to
avoid any locking issues.Signed-off-by: James Bottomley
09 Sep, 2005
4 commits
-
From: Alan Stern
This patch (as559b) adds a new routine, scsi_unprep_request, which
gets called every place a request is requeued. (That includes
scsi_queue_insert as well as scsi_requeue_command.) It also changes
scsi_kill_requests to make it call __scsi_done with result equal to
DID_NO_CONNECT << 16. (I'm not sure if it's necessary to call
scsi_init_cmd_errh here; maybe you can check on that.) Finally, the
patch changes the return value from scsi_end_request, to avoid
returning a stale pointer in the case where the request was requeued.
Fortunately the return value is used in only place, and the change
actually simplified it.Signed-off-by: Alan Stern
Rejections fixed up and
Signed-off-by: James Bottomley -
If a filesystem, while writing out data, decides that it is good
to issue a cache flush on a SCSI drive (or other 'sd' device), it will
call blkdev_issue_flush which calls ->issue_flush_fn which is
scsi_issue_flush_fn.
This calls sd_issue_flush which calls sd_sync_cache, which calls
scsi_execute_request.
This will (as sshdr != NULL) call
kmalloc(SCSI_SENSE_BUFFERSIZE, GFP_KERNEL)If memory is tight, the presence of GFP_KERNEL may cause write
requests to be sent to some filesystem to free up memory, however if
that filesystem is waiting for the issue_flush_fn to complete, you
could get a deadlock.I wonder if it might be more appropriate to use GFP_NOIO as in the
following patch.I wonder if it might be even more appropriate to cope better with a
kmalloc failure, especially as in this use, sd_sync_cache only will
use the sense information to print out a more informative error
message.Signed-off-by: Neil Brown
Signed-off-by: James Bottomley -
This patch (as544) adds a private entry point to scsi_remove_device, for
use when callers already own the scan_mutex. The appropriate callers are
modified to use the new entry point.Signed-off-by: Alan Stern
Signed-off-by: James Bottomley -
This patch (as543) adds a private entry point to scsi_scan_target, for use
when the caller already owns the scan_mutex, and updates the kerneldoc for
that routine (which was badly out-of-date). It converts scsi_scan_channel
to use the new entry point. Lastly, it modifies scsi_get_host_dev to make
it acquire the scan_mutex, necessary since the routine adds a new
scsi_device even if it doesn't do any actual scanning.Signed-off-by: Alan Stern
Signed-off-by: James Bottomley
08 Sep, 2005
32 commits
-
With the previous commit that introduces the klist enhancements, we can
now re-do 2b7d6a8cb9718fc1d9e826201b64909c44a915f4 again. -
The problem is that klists claim to provide semantics for safe traversal of
lists which are being modified. The failure case is when traversal of a
list causes element removal (a fairly common case). The issue is that
although the list node is refcounted, if it is embedded in an object (which
is universally the case), then the object will be freed regardless of the
klist refcount leading to slab corruption because the klist iterator refers
to the prior element to get the next.The solution is to make the klist take and release references to the
embedding object meaning that the embedding object won't be released until
the list relinquishes the reference to it.(akpm: fast-track this because it's needed for the 2.6.13 scsi merge)
Signed-off-by: James Bottomley
Signed-off-by: Greg Kroah-Hartman
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
Revert commit 2b7d6a8cb9718fc1d9e826201b64909c44a915f4.
The "fix" was known to not even compile. Duh. That's not a fix.
That's just stupid.Signed-off-by: Linus Torvalds
-
Signed-off-by: Al Viro
Signed-off-by: Linus Torvalds -
Now that asm-powerpc/* is using ifdefs on __powerpc64__ we need to add it
to CHECKFLAGS on ppc64.Signed-off-by: Al Viro
Signed-off-by: Linus Torvalds -
Signed-off-by: Al Viro
Signed-off-by: Linus Torvalds -
Signed-off-by: Al Viro
Signed-off-by: Linus Torvalds -
Signed-off-by: Al Viro
Signed-off-by: Linus Torvalds -
CARD_... in hisax are all used with #if; CARD_FN_ENTERNOW_PCI lacks define
to 0 if corresponding config option is not set.Signed-off-by: Al Viro
Signed-off-by: Linus Torvalds -
NDEBUG and NDEBUG_ABORT are almost always used as integers in NCR5380; added
define to 0 if they are not defined, switched lone ifdef NDEBUG into if.Signed-off-by: Al Viro
Signed-off-by: Linus Torvalds -
All uses of ADDRLEN are comparisons with 64 (it's an address width).
added define to 32 (again, we only care about comparisons with 64)
if not defined.Signed-off-by: Al Viro
Signed-off-by: Linus Torvalds -
elf_aux is userland code; it uses symbol (ELF_CLASS) that doesn't exist in
userland headers; pulled into kernel-offsets.h, switched elf_aux to using it.Signed-off-by: Al Viro
Signed-off-by: Linus Torvalds -
parcelfarce is dead...
Signed-off-by: Al Viro
Signed-off-by: Linus Torvalds -
Signed-off-by: Al Viro
Signed-off-by: Linus Torvalds -
A piece of the UML stubs patch got lost - it has
Killed STUBS_CFLAGS - it's not needed and the only remaining use had been
gratitious - it only polluted CFLAGS
in description and does remove it in arch/um/Makefile-x86_64, but forgets to
do the same in i386 counterpart. Lost chunk follows:Signed-off-by: Al Viro
Signed-off-by: Linus Torvalds -
Sanitized and fixed floppy dependencies: split the messy dependencies for
BLK_DEV_FD by introducing a new symbol (ARCH_MAY_HAVE_PC_FDC), making
BLK_DEV_FD depend on that one and taking declarations of ARCH_MAY_HAVE_PC_FDC
to arch/*/Kconfig. While we are at it, fixed several obvious cases when
BLK_DEV_FD should have been excluded (architectures lacking asm/floppy.h
are *not* going to have floppy.c compile, let alone work).If you can come up with better name for that ("this architecture might
have working PC-compatible floppy disk controller"), you are more than
welcome - just s/ARCH_MAY_HAVE_PC_FDC/your_prefered_name/g in the patch
below...Signed-off-by: Al Viro
Signed-off-by: Linus Torvalds -
u64 is not uintptr_t; unsigned long is...
Signed-off-by: Al Viro
Signed-off-by: Linus Torvalds -
Most of the patch is whitespace cleanup, but more importantly, this patch
checks to see whether a callback is set before calling it. On cx88 boards
(currently the only boards using lgdt330x in 2.6.13) every callback is set.
However, newer drivers currently in development leave a callback undefined,
and lgdt330x must not call it if it isn't defined.Signed-off-by: Patrick Boettcher
Signed-off-by: Michael Krufky
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
Patrick Keene wrote to the linux-dvb list, asking where in menuconfig he
can enable dvb-bt8xx for his AVerMedia DVB card. I pointed the following
out to him:config DVB_BT8XX
tristate "Nebula/Pinnacle PCTV/Twinhan PCI cards"It has been agreed upon that this description is extremely misleading.
This patch changes the one-liner description text of dvb-bt8xx to something
more meaningful, and adds AVerMedia to the detailed description.Signed-off-by: Michael Krufky
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
Rewrite of the Indycam / VINO video v4l2 drivers for the SGI Indy.
Signed-off-by: Ralf Baechle
Signed-off-by: Mikael Nousiainen
Cc:
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
Fix http://bugzilla.kernel.org/show_bug.cgi?id=4857
When pivot_root is called from an init script in an initramfs environment,
it causes a circular reference in the mount tree.The cause of this is that pivot_root() is not prepared to handle pivoting
an unattached mount. In an initramfs environment, rootfs is the root of
the namespace, and so it is not attached.This patch fixes this and related problems, by returning -EINVAL if either
the current root or the new root is detached.Signed-off-by: Miklos Szeredi
Acked-by: Al Viro
Cc:
Cc: Christoph Hellwig
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
This patch fixes a race condition where in system used to hang or sometime
crash within minutes when kprobes are inserted on ISR routine and a task
routine.The fix has been stress tested on i386, ia64, pp64 and on x86_64. To
reproduce the problem insert kprobes on schedule() and do_IRQ() functions
and you should see hang or system crash.Signed-off-by: Anil S Keshavamurthy
Signed-off-by: Ananth N Mavinakayanahalli
Acked-by: Prasanna S Panchamukhi
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds
