dev_base_lock is the legacy way to lock the device list, and is planned
to disappear. (writers hold RTNL, readers hold RCU lock)
Convert aoecmd_cfg_pkts() to RCU locking.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Cc: "Ed L. Cashin" <ecashin@coraid.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
percpu.h is included by sched.h and module.h and thus ends up being
included when building most .c files. percpu.h includes slab.h which
in turn includes gfp.h making everything defined by the two files
universally available and complicating inclusion dependencies.
percpu.h -> slab.h dependency is about to be removed. Prepare for
this change by updating users of gfp and slab facilities include those
headers directly instead of assuming availability. As this conversion
needs to touch large number of source files, the following script is
used as the basis of conversion.
http://userweb.kernel.org/~tj/misc/slabh-sweep.py
The script does the followings.
* Scan files for gfp and slab usages and update includes such that
only the necessary includes are there. ie. if only gfp is used,
gfp.h, if slab is used, slab.h.
* When the script inserts a new include, it looks at the include
blocks and try to put the new include such that its order conforms
to its surrounding. It's put in the include block which contains
core kernel includes, in the same order that the rest are ordered -
alphabetical, Christmas tree, rev-Xmas-tree or at the end if there
doesn't seem to be any matching order.
* If the script can't find a place to put a new include (mostly
because the file doesn't have fitting include block), it prints out
an error message indicating which .h file needs to be added to the
file.
The conversion was done in the following steps.
1. The initial automatic conversion of all .c files updated slightly
over 4000 files, deleting around 700 includes and adding ~480 gfp.h
and ~3000 slab.h inclusions. The script emitted errors for ~400
files.
2. Each error was manually checked. Some didn't need the inclusion,
some needed manual addition while adding it to implementation .h or
embedding .c file was more appropriate for others. This step added
inclusions to around 150 files.
3. The script was run again and the output was compared to the edits
from #2 to make sure no file was left behind.
4. Several build tests were done and a couple of problems were fixed.
e.g. lib/decompress_*.c used malloc/free() wrappers around slab
APIs requiring slab.h to be added manually.
5. The script was run on all .h files but without automatically
editing them as sprinkling gfp.h and slab.h inclusions around .h
files could easily lead to inclusion dependency hell. Most gfp.h
inclusion directives were ignored as stuff from gfp.h was usually
wildly available and often used in preprocessor macros. Each
slab.h inclusion directive was examined and added manually as
necessary.
6. percpu.h was updated not to include slab.h.
7. Build test were done on the following configurations and failures
were fixed. CONFIG_GCOV_KERNEL was turned off for all tests (as my
distributed build env didn't work with gcov compiles) and a few
more options had to be turned off depending on archs to make things
build (like ipr on powerpc/64 which failed due to missing writeq).
* x86 and x86_64 UP and SMP allmodconfig and a custom test config.
* powerpc and powerpc64 SMP allmodconfig
* sparc and sparc64 SMP allmodconfig
* ia64 SMP allmodconfig
* s390 SMP allmodconfig
* alpha SMP allmodconfig
* um on x86_64 SMP allmodconfig
8. percpu.h modifications were reverted so that it could be applied as
a separate patch and serve as bisection point.
Given the fact that I had only a couple of failures from tests on step
6, I'm fairly confident about the coverage of this conversion patch.
If there is a breakage, it's likely to be something in one of the arch
headers which should be easily discoverable easily on most builds of
the specific arch.
Signed-off-by: Tejun Heo <tj@kernel.org>
Guess-its-ok-by: Christoph Lameter <cl@linux-foundation.org>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>
Prevent the AoE block driver from creating cache aliases of page cache
pages on machines with virtually indexed caches.
Building kernels on an AT91SAM9G20 board without this patch fails with
segmentation faults after a couple of passes.
Signed-off-by: Peter Horton <zero@colonel-panic.org>
Cc: "Ed L. Cashin" <ecashin@coraid.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* Use ATA_CMD_* defines instead of WIN_* ones.
* Include <linux/ata.h> directly instead of through <linux/hdreg.h>.
Cc: Ed L. Cashin <ecashin@coraid.com>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next-2.6: (1075 commits)
myri10ge: update driver version number to 1.4.3-1.369
r8169: add shutdown handler
r8169: preliminary 8168d support
r8169: support additional 8168cp chipset
r8169: change default behavior for mildly identified 8168c chipsets
r8169: add a new 8168cp flavor
r8169: add a new 8168c flavor (bis)
r8169: add a new 8168c flavor
r8169: sync existing 8168 device hardware start sequences with vendor driver
r8169: 8168b Tx performance tweak
r8169: make room for more specific 8168 hardware start procedure
r8169: shuffle some registers handling around (8168 operation only)
r8169: new phy init parameters for the 8168b
r8169: update phy init parameters
r8169: wake up the PHY of the 8168
af_key: fix SADB_X_SPDDELETE response
ath9k: Fix return code when ath9k_hw_setpower() fails on reset
ath9k: remove nasty FAIL macro from ath9k_hw_reset()
gre: minor cleanups in netlink interface
gre: fix copy and paste error
...
Move stats related fields - stamp, in_flight, dkstats - from disk to
part0 and unify stat handling such that...
* part_stat_*() now updates part0 together if the specified partition
is not part0. ie. part_stat_*() are now essentially all_stat_*().
* {disk|all}_stat_*() are gone.
* part_round_stats() is updated similary. It handles part0 stats
automatically and disk_round_stats() is killed.
* part_{inc|dec}_in_fligh() is implemented which automatically updates
part0 stats for parts other than part0.
* disk_map_sector_rcu() is updated to return part0 if no part matches.
Combined with the above changes, this makes NULL special case
handling in callers unnecessary.
* Separate stats show code paths for disk are collapsed into part
stats show code paths.
* Rename disk_stat_lock/unlock() to part_stat_lock/unlock()
While at it, reposition stat handling macros a bit and add missing
parentheses around macro parameters.
Signed-off-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
Move disk->capacity to part0->nr_sects and convert all users who
directly accessed the field to use {get|set}_capacity(). This is done
early to allow the __dev field to be moved.
Signed-off-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
There are two variants of stat functions - ones prefixed with double
underbars which don't care about preemption and ones without which
disable preemption before manipulating per-cpu counters. It's unclear
whether the underbarred ones assume that preemtion is disabled on
entry as some callers don't do that.
This patch unifies diskstats access by implementing disk_stat_lock()
and disk_stat_unlock() which take care of both RCU (for partition
access) and preemption (for per-cpu counter access). diskstats access
should always be enclosed between the two functions. As such, there's
no need for the versions which disables preemption. They're removed
and double underbars ones are renamed to drop the underbars. As an
extra argument is added, there's no danger of using the old version
unconverted.
disk_stat_lock() uses get_cpu() and returns the cpu index and all
diskstat functions which access per-cpu counters now has @cpu
argument to help RT.
This change adds RCU or preemption operations at some places but also
collapses several preemption ops into one at others. Overall, the
performance difference should be negligible as all involved ops are
very lightweight per-cpu ones.
Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
disk->part[] is protected by its matching bdev's lock. However,
non-critical accesses like collecting stats and printing out sysfs and
proc information used to be performed without any locking. As
partitions can come and go dynamically, partitions can go away
underneath those non-critical accesses. As some of those accesses are
writes, this theoretically can lead to silent corruption.
This patch fixes the race by using RCU for the partition array and dev
reference counter to hold partitions.
* Rename disk->part[] to disk->__part[] to make sure no one outside
genhd layer proper accesses it directly.
* Use RCU for disk->__part[] dereferencing.
* Implement disk_{get|put}_part() which can be used to get and put
partitions from gendisk respectively.
* Iterators are implemented to help iterate through all partitions
safely.
* Functions which require RCU readlock are marked with _rcu suffix.
* Use disk_put_part() in __blkdev_put() instead of directly putting
the contained kobject.
Signed-off-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
This patch makes the following misc updates in preparation for
disk->part dereference fix and extended block devt support.
* implment part_to_disk()
* fix comment about gendisk->part indexing
* rename get_part() to disk_map_sector()
* don't use n which is always zero while printing disk information in
diskstats_show()
Signed-off-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
get_part() is fairly expensive, as it O(N) loops over partitions
to find the right one. In lots of normal IO paths we end up looking
up the partition twice, to make matters even worse. Change the
stat add code to accept a passed in partition instead.
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
Andrew Morton pointed out that the "too many targets" message in patch 2 could
be printed for failing GFP_ATOMIC allocations. This patch makes the messages
more specific.
Signed-off-by: Ed L. Cashin <ecashin@coraid.com>
Cc: Greg KH <greg@kroah.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
An AoE target provides an estimate of the number of outstanding commands that
the AoE initiator can send before getting a response. The aoe_maxout
parameter provides a way to set an even lower limit. It will not allow a user
to use more outstanding commands than the target permits. If a user discovers
a problem with a large setting, this parameter provides a way for us to work
with them to debug the problem. We expect to improve the dynamic window
sizing algorithm and drop this parameter. For the time being, it is a
debugging aid.
Signed-off-by: Ed L. Cashin <ecashin@coraid.com>
Cc: Greg KH <greg@kroah.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
An aoe driver user who had about 70 AoE targets found that he was hitting a
BUG in sysfs_create_file because the aoe driver was trying to tell the kernel
about an AoE device more than once. Each AoE device was reachable by several
local network interfaces, and multiple ATA device indentify responses were
returning from that single device.
This patch eliminates a race condition so that aoe always informs the block
layer of a new AoE device once in the presence of multiple incoming ATA device
identify responses.
Signed-off-by: Ed L. Cashin <ecashin@coraid.com>
Cc: Greg KH <greg@kroah.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
What this Patch Does
Even before this recent series of 12 patches to 2.6.22-rc4, the aoe
driver was reusing a small set of skbs that were allocated once and
were only used for outbound AoE commands.
The network layer cannot be allowed to put_page on the data that is
still associated with a bio we haven't returned to the block layer,
so the aoe driver (even before the patch under discussion) is still
the owner of skbs that have been handed to the network layer for
transmission. We need to keep track of these skbs so that we can
free them, but by tracking them, we can also easily re-use them.
The new patch was a response to the behavior of certain network
drivers. We cannot reuse an skb that the network driver still has
in its transmit ring. Network drivers can defer transmit ring
cleanup and then use the state in the skb to determine how many data
segments to clean up in its transmit ring. The tg3 driver is one
driver that behaves in this way.
When the network driver defers cleanup of its transmit ring, the aoe
driver can find itself in a situation where it would like to send an
AoE command, and the AoE target is ready for more work, but the
network driver still has all of the pre-allocated skbs. In that
case, the new patch just calls alloc_skb, as you'd expect.
We don't want to get carried away, though. We try not to do
excessive allocation in the write path, so we cap the number of skbs
we dynamically allocate.
Probably calling it a "dynamic pool" is misleading. We were already
trying to use a small fixed-size set of pre-allocated skbs before
this patch, and this patch just provides a little headroom (with a
ceiling, though) to accomodate network drivers that hang onto skbs,
by allocating when needed. The d->skbpool_hd list of allocated skbs
is necessary so that we can free them later.
We didn't notice the need for this headroom until AoE targets got
fast enough.
Alternatives
If the network layer never did a put_page on the pages in the bio's
we get from the block layer, then it would be possible for us to
hand skbs to the network layer and forget about them, allowing the
network layer to free skbs itself (and thereby calling our own
skb->destructor callback function if we needed that). In that case
we could get rid of the pre-allocated skbs and also the
d->skbpool_hd, instead just calling alloc_skb every time we wanted
to transmit a packet. The slab allocator would effectively maintain
the list of skbs.
Besides a loss of CPU cache locality, the main concern with that
approach the danger that it would increase the likelihood of
deadlock when VM is trying to free pages by writing dirty data from
the page cache through the aoe driver out to persistent storage on
an AoE device. Right now we have a situation where we have
pre-allocation that corresponds to how much we use, which seems
ideal.
Of course, there's still the separate issue of receiving the packets
that tell us that a write has successfully completed on the AoE
target. When memory is low and VM is using AoE to flush dirty data
to free up pages, it would be perfect if there were a way for us to
register a fast callback that could recognize write command
completion responses. But I don't think the current problems with
the receive side of the situation are a justification for
exacerbating the problem on the transmit side.
Signed-off-by: Ed L. Cashin <ecashin@coraid.com>
Cc: Greg KH <greg@kroah.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Eric Dumazet
Tejun Heo
Andrew Morton
Peter Horton
David S. Miller
Bartlomiej Zolnierkiewicz
Harvey Harrison
Linus Torvalds
Jens Axboe
Linus Torvalds
Ed L. Cashin