mthca: initialize send and receive queue locks separately
lockdep identifies a lock by the call site of its initialization. By
initializing the send and receive queue locks in mthca_wq_init() we confuse
lockdep. It warns that that the ordered acquiry of both locks in
mthca_modify_qp() is recursive acquiry of one lock:
=============================================
[ INFO: possible recursive locking detected ]
---------------------------------------------
modprobe/1192 is trying to acquire lock:
(&wq->lock){....}, at: [<f892b4db>] mthca_modify_qp+0x60/0xa7b [ib_mthca]
but task is already holding lock:
(&wq->lock){....}, at: [<f892b4ce>] mthca_modify_qp+0x53/0xa7b [ib_mthca]
Initializing the locks separately in mthca_alloc_qp_common() stops the
warning and will let lockdep enforce proper ordering on paths that acquire
both locks.
Signed-off-by: Zach Brown <[email protected]>
Cc: Roland Dreier <[email protected]>
Cc: Arjan van de Ven <[email protected]>
Cc: Ingo Molnar <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
AOP_TRUNCATED_PAGE victims in read_pages() belong in the LRU
Nick Piggin rightly pointed out that the introduction of AOP_TRUNCATED_PAGE
to read_pages() was wrong to leave A_T_P victim pages in the page cache but
not put them in the LRU. Failing to do so hid them from the VM.
A_T_P just means that the aop method unlocked the page rather than
performing IO. It would be very rare that the page was truncated between
the unlock and testing A_T_P. So we leave the pages in the LRU for likely
reuse soon rather than backing them back out of the page cache. We do this
by matching the behaviour before the A_T_P introduction which added pages
to the LRU regardless of what ->readpage() did.
This doesn't include the unrelated cleanup in Nick's initial fix which
changed read_pages() to return void to match its only caller's behaviour of
ignoring errors.
Signed-off-by: Nick Piggin <[email protected]>
Signed-off-by: Zach Brown <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
The binrpm-pkg target uses KERNELRELEASE when generated its .spec file.
When binrpm-pkg was the first build target run in a tree it generated the
.spec before kernel.release so the Version: tag in the .spec was empty.
I don't know if this is the best fix, but binrpm-pkg works when we
explicitly build kernel.release before descending into package-dir.
Signed-off-by: Zach Brown <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Sam Ravnborg <[email protected]>
If qla2x00_probe_one() fails before calling request_irq() but gets to
qla2x00_free_device() then it will mistakenly try to free an irq it didn't
request. It's chosing to free based on ha->pdev->irq which is always set.
host->irq is set after request_irq() succeeds so let's use that to decide
to free or not.
This was observed and tested when a silly set of circumstances lead to
firmware loading failing on a 2100.
Signed-off-by: Zach Brown <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: James Bottomley <[email protected]>
During other work I noticed that ip_append_data() seemed to be forgetting to
include the frag gap in its calculation of a fragment that consumes the rest of
the payload. Herbert confirmed that this was a bug that snuck in during a
previous rework.
Signed-off-by: Zach Brown <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
This closes a race where an ipq6hashfn() caller could get a hash value
and race with the cycling of the random seed. By the time they got to
the read_lock they'd have a stale hash value and might not find
previous fragments of their datagram.
This matches the previous patch to IPv4.
Signed-off-by: Zach Brown <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
This matches the fix for a bug seen on x86-64. Test booted on old hardware
that had 32 byte cachelines to begin with.
Signed-off-by: Zach Brown <[email protected]>
Signed-off-by: Sam Ravnborg <[email protected]>
Align the start of the per-cpu section to the configured number of bytes in a
cache line. This stops a BUG_ON() from triggering in load_module() when
DEFINE_PER_CPU() is used in a module and the section isn't cacheline-aligned.
Rusty also found this and sent a patch in a while ago
(http://lkml.org/lkml/2004/10/19/17), I don't know what came of that.
Signed-off-by: Zach Brown <[email protected]>
Signed-off-by: Andi Kleen <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
I noticed that list.h init functions were evaluating macro arguments
multiple times and thought it might be nice to protect the unsuspecting
caller. Converting the macros to inline functions seems to reduce code
size, too. A i386 defconfig build with gcc 3.3.3 from fc4:
text data bss dec hex filename
3573148 565664 188828 4327640 4208d8 vmlinux.before
3572177 565664 188828 4326669 42050d vmlinux
add/remove: 0/0 grow/shrink: 11/144 up/down: 88/-1016 (-928)
There was no difference in checkstack output.
Signed-off-by: Zach Brown <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
Node messaging via tcp. Used by the dlm and the file system for point
to point communication between nodes.
Signed-off-by: Mark Fasheh <[email protected]>
Signed-off-by: Kurt Hackel <[email protected]>
Very simple printk wrapper which adds the ability to enable various
sets of debug messages at run-time.
Signed-off-by: Mark Fasheh <[email protected]>
Signed-off-by: Kurt Hackel <[email protected]>
readpage(), prepare_write(), and commit_write() callers are updated to
understand the special return code AOP_TRUNCATED_PAGE in the style of
writepage() and WRITEPAGE_ACTIVATE. AOP_TRUNCATED_PAGE tells the caller that
the callee has unlocked the page and that the operation should be tried again
with a new page. OCFS2 uses this to detect and work around a lock inversion in
its aop methods. There should be no change in behaviour for methods that don't
return AOP_TRUNCATED_PAGE.
WRITEPAGE_ACTIVATE is also prepended with AOP_ for consistency and they are
made enums so that kerneldoc can be used to document their semantics.
Signed-off-by: Zach Brown <[email protected]>
put_ioctx's refcount debugging was doing an atomic_read after dropping its
reference when it wasn't the last ref, leaving a tiny race for another freeing
thread to sneak into. This shifts the debugging before the ops, uses BUG_ON,
and reformats the defines a little. Sadly, moving to inlines increased the
code size but this change decreases the code size by a whole 9 bytes :)
Signed-off-by: Zach Brown <[email protected]>
Cc: Benjamin LaHaise <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
Sync iocbs have a life cycle that don't need a kioctx. Their retrying, if
any, is done in the context of their owner who has allocated them on the
stack.
The sole user of a sync iocb's ctx reference was aio_complete() checking for
an elevated iocb ref count that could never happen. No path which grabs an
iocb ref has access to sync iocbs.
If we were to implement sync iocb cancelation it would be done by the owner of
the iocb using its on-stack reference.
Removing this chunk from aio_complete allows us to remove the entire kioctx
instance from mm_struct, reducing its size by a third. On a i386 testing box
the slab size went from 768 to 504 bytes and from 5 to 8 per page.
Signed-off-by: Zach Brown <[email protected]>
Acked-by: Benjamin LaHaise <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
AIO was adding a new context's max requests to the global total before
testing if that resulting total was over the global limit. This let
innocent tasks get their new limit tested along with a racing guilty task
that was crossing the limit. This serializes the _nr accounting with a
spinlock It also switches to using unsigned long for the global totals.
Individual contexts are still limited to an unsigned int's worth of
requests by the syscall interface.
The problem and fix were verified with a simple program that spun creating
and destroying a context while holding on to another long lived context.
Before the patch a task creating a tiny context could get a spurious EAGAIN
if it raced with a task creating a very large context that overran the
limit.
Signed-off-by: Zach Brown <[email protected]>
Cc: Benjamin LaHaise <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
lock_kiocb() was introduced to serialize retrying and cancellation. In the
process of doing so it tried to sleep waiting for KIF_LOCKED while holding
the ctx_lock spinlock. Recent fixes have ensured that multiple concurrent
retries won't be attempted for a given iocb. Cancel has other problems and
has no significant in-tree users that have been complaining about it. So
for the immediate future we'll revert sleeping with the lock held and will
address proper cancellation and retry serialization in the future.
Signed-off-by: Zach Brown <[email protected]>
Acked-by: Benjamin LaHaise <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
Recently aio_p{read,write} changed to perform retries internally rather
than returning -EIOCBRETRY. This inadvertantly resulted in always calling
aio_{read,write} with ki_left at 0 which would in turn immediately return
0. Harmless, but we can avoid this call by checking in the caller.
Signed-off-by: Zach Brown <[email protected]>
Signed-off-by: Benjamin LaHaise <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
Only one of the run or kick path is supposed to put an iocb on the run
list. If both of them do it than one of them can end up referencing a
freed iocb. The kick path could delete the task_list item from the wait
queue before getting the ctx_lock and putting the iocb on the run list.
The run path was testing the task_list item outside the lock so that it
could catch ki_retry methods that return -EIOCBRETRY *without* putting the
iocb on a wait queue and promising to call kick_iocb. This unlocked check
could then race with the kick path to cause both to try and put the iocb on
the run list.
The patch stops the run path from testing task_list by requring that any
ki_retry that returns -EIOCBRETRY *must* guarantee that kick_iocb() will be
called in the future. aio_p{read,write}, the only in-tree -EIOCBRETRY
users, are updated.
Signed-off-by: Zach Brown <[email protected]>
Signed-off-by: Benjamin LaHaise <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
Only one of the run or kick path is supposed to put an iocb on the run
list. If both of them do it than one of them can end up referencing a
freed iocb. The kick patch could set the Kicked bit before acquiring the
ctx_lock and putting the iocb on the run list. The run path, while holding
the ctx_lock, could see this partial kick and mistake it for a kick that
was deferred while it was doing work with the run_list NULLed out. It
would then race with the kick thread to add the iocb to the run list.
This patch moves the kick setting under the ctx_lock so that only one of
the kick or run path queues the iocb on the run list, as intended.
Signed-off-by: Zach Brown <[email protected]>
Signed-off-by: Benjamin LaHaise <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
Add smp_mb__after_clear_bit() to unlock_kiocb()
AIO's use of wait_on_bit_lock()/wake_up_bit() forgot to add a barrier
between clearing its lock bit and calling wake_up_bit() so wake_up_bit()'s
unlocked waitqueue_active() can race. This puts AIO's use in line with the
others and the comment above wake_up_bit().
Signed-off-by: Zach Brown <[email protected]>
Acked-by: Benjamin LaHaise <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
Dunno if there was a conscious decision to leave it out, but if you're
happy with adding some help text for it here's a patch against 2.6.13-mm1..
Signed-off-by: Sam Ravnborg <[email protected]>
Zach Brown