15 Apr, 2009
11 commits
-
If we have processes that are working in close proximity to each
other on disk, we don't want to idle wait. Instead allow the close
process to issue a request, getting better aggregate bandwidth.
The anticipatory scheduler has similar checks, noop and deadline do
not need it since they don't care about process io mappings.The code for CFQ is a little more involved though, since we split
request queues into per-process contexts.This fixes a performance problem with eg dump(8), since it uses
several processes in some silly attempt to speed IO up. Even if
dump(8) isn't really a valid case (it should be fixed by using
CLONE_IO), there are other cases where we see close processes
and where idling ends up hurting performance.Credit goes to Jeff Moyer for writing the
initial implementation.Signed-off-by: Jens Axboe
-
Makes it easier to read the traces.
Signed-off-by: Jens Axboe
-
We only kick the dispatch for an idling queue, if we think it's a
(somewhat) fully merged request. Also allow a kick if we have other
busy queues in the system, since we don't want to risk waiting for
a potential merge in that case. It's better to get some work done and
proceed.Signed-off-by: Jens Axboe
-
It's called from the workqueue handlers from process context, so
we always have irqs enabled when entered.Signed-off-by: Jens Axboe
-
Remove code handling bio_alloc failure with __GFP_WAIT.
GFP_KERNEL implies __GFP_WAIT.Signed-off-by: Nikanth Karthikesan
Signed-off-by: Jens Axboe -
blk_rq_unmap_user() returns -EFAULT if a program passes an invalid
address to kernel. SG_IO path needs to pass the returned value to user
space instead of ignoring it.Signed-off-by: FUJITA Tomonori
Signed-off-by: Jens Axboe -
"Zhang, Yanmin" reports that commit
b029195dda0129b427c6e579a3bb3ae752da3a93 introduced a regression
of about 50% with sequential threaded read workloads. The test
case is:tiotest -k0 -k1 -k3 -f 80 -t 32
which starts 32 threads each reading a 80MB file. Twiddle the kick
queue logic so that we do start IO immediately, if it appears to be
a fully merged request. We can't really detect that, so just check
if the request is bigger than a page or not. The assumption is that
since single bio issues will first queue a single request with just
one page attached and then later do merges on that, if we already
have more than a page worth of data in the request, then the request
is most likely good to go.Verified that this doesn't cause a regression with the test case that
commit b029195dda0129b427c6e579a3bb3ae752da3a93 was fixing. It does not,
we still see maximum sized requests for the queue-then-merge cases.Signed-off-by: Jens Axboe
-
We can just use the block layer BLK_RW_SYNC/ASYNC defines now.
Signed-off-by: Jens Axboe
-
We can just use the block layer BLK_RW_SYNC/ASYNC defines now.
Signed-off-by: Jens Axboe
-
Signed-off-by: Jens Axboe
-
Credit goes to Andrew Morton for spotting this one.
Signed-off-by: Jens Axboe
08 Apr, 2009
2 commits
-
…nel/git/tip/linux-2.6-tip
* 'tracing-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
branch tracer, intel-iommu: fix build with CONFIG_BRANCH_TRACER=y
branch tracer: Fix for enabling branch profiling makes sparse unusable
ftrace: Correct a text align for event format output
Update /debug/tracing/README
tracing/ftrace: alloc the started cpumask for the trace file
tracing, x86: remove duplicated #include
ftrace: Add check of sched_stopped for probe_sched_wakeup
function-graph: add proper initialization for init task
tracing/ftrace: fix missing include string.h
tracing: fix incorrect return type of ns2usecs()
tracing: remove CALLER_ADDR2 from wakeup tracer
blktrace: fix pdu_len when tracing packet command requests
blktrace: small cleanup in blk_msg_write()
blktrace: NUL-terminate user space messages
tracing: move scripts/trace/power.pl to scripts/tracing/power.pl -
* 'for-linus' of git://git.kernel.dk/linux-2.6-block:
loop: mutex already unlocked in loop_clr_fd()
cfq-iosched: don't let idling interfere with plugging
block: remove unused REQ_UNPLUG
cfq-iosched: kill two unused cfqq flags
cfq-iosched: change dispatch logic to deal with single requests at the time
mflash: initial support
cciss: change to discover first memory BAR
cciss: kernel scan thread for MSA2012
cciss: fix residual count for block pc requests
block: fix inconsistency in I/O stat accounting code
block: elevator quiescing helpers
07 Apr, 2009
8 commits
-
* 'upstream-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jgarzik/libata-dev:
sata_mv: shorten register names
sata_mv: workaround errata SATA#13
sata_mv: cosmetic renames
sata_mv: workaround errata SATA#26
sata_mv: workaround errata PCI#7
sata_mv: replace 0x1f with ATA_PIO4 (v2)
sata_mv: fix irq mask races
sata_mv: revert SoC irq breakage
libata: ahci enclosure management bios workaround
ata: Add TRIM infrastructure
ata_piix: VGN-BX297XP wants the controller power up on suspend
libata: Remove some redundant casts from pata_octeon_cf.c
pata_artop: typo -
When CFQ is waiting for a new request from a process, currently it'll
immediately restart queuing when it sees such a request. This doesn't
work very well with streamed IO, since we then end up splitting IO
that would otherwise have been merged nicely. For a simple dd test,
this causes 10x as many requests to be issued as we should have.
Normally this goes unnoticed due to the low overhead of requests
at the device side, but some hardware is very sensitive to request
sizes and there it can cause big slow downs.Signed-off-by: Jens Axboe
-
The request inherits the unplug flag from the bio, but it isn't actually
used. The bio flag stops at __make_request(), which tells it to unplug
after submission. Passing it on to the request doesn't make any sense.Signed-off-by: Jens Axboe
-
We only manipulate the must_dispatch and queue_new flags, they are not
tested anymore. So get rid of them.Signed-off-by: Jens Axboe
-
The IO scheduler core calls into the IO scheduler dispatch_request hook
to move requests from the IO scheduler and into the driver dispatch
list. It only does so when the dispatch list is empty. CFQ moves several
requests to the dispatch list, which can cause higher latencies if we
suddenly have to switch to some important sync IO. Change the logic to
move one request at the time instead.This should almost be functionally equivalent to what we did before,
except that we now honor 'quantum' as the maximum queue depth at the
device side from any single cfqq. If there's just a single active
cfqq, we allow up to 4 times the normal quantum.Signed-off-by: Jens Axboe
-
This forces in_flight to be zero when turning off or on the I/O stat
accounting and stops updating I/O stats in attempt_merge() when
accounting is turned off.Signed-off-by: Jerome Marchand
Signed-off-by: Jens Axboe -
Simple helper functions to quiesce the request queue. These are
currently only used for switching IO schedulers on-the-fly, but
we can use them to properly switch IO accounting on and off as well.Signed-off-by: Jerome Marchand
Signed-off-by: Jens Axboe -
Fix a typo (this was in the original patch but was not merged when the code
fixes were for some reason)Signed-off-by: Alan Cox
Signed-off-by: Jeff Garzik
06 Apr, 2009
5 commits
-
By default, CFQ will anticipate more IO from a given io context if the
previously completed IO was sync. This used to be fine, since the only
sync IO was reads and O_DIRECT writes. But with more "normal" sync writes
being used now, we don't want to anticipate for those.Add a bio/request flag that informs the IO scheduler that this is a sync
request that we should not idle for. Introduce WRITE_ODIRECT specifically
for O_DIRECT writes, and make sure that the other sync writes set this
flag.Signed-off-by: Jens Axboe
Signed-off-by: Linus Torvalds -
For the older SSD devices that don't do command queuing, we do want to
enable plugging to get better merging.Signed-off-by: Jens Axboe
Signed-off-by: Linus Torvalds -
This makes sure that we never wait on async IO for sync requests, instead
of doing the split on writes vs reads.Signed-off-by: Jens Axboe
Signed-off-by: Linus Torvalds -
* 'tracing-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip: (413 commits)
tracing, net: fix net tree and tracing tree merge interaction
tracing, powerpc: fix powerpc tree and tracing tree interaction
ring-buffer: do not remove reader page from list on ring buffer free
function-graph: allow unregistering twice
trace: make argument 'mem' of trace_seq_putmem() const
tracing: add missing 'extern' keywords to trace_output.h
tracing: provide trace_seq_reserve()
blktrace: print out BLK_TN_MESSAGE properly
blktrace: extract duplidate code
blktrace: fix memory leak when freeing struct blk_io_trace
blktrace: fix blk_probes_ref chaos
blktrace: make classic output more classic
blktrace: fix off-by-one bug
blktrace: fix the original blktrace
blktrace: fix a race when creating blk_tree_root in debugfs
blktrace: fix timestamp in binary output
tracing, Text Edit Lock: cleanup
tracing: filter fix for TRACE_EVENT_FORMAT events
ftrace: Using FTRACE_WARN_ON() to check "freed record" in ftrace_release()
x86: kretprobe-booster interrupt emulation code fix
...Fix up trivial conflicts in
arch/parisc/include/asm/ftrace.h
include/linux/memory.h
kernel/extable.c
kernel/module.c -
* git://git.kernel.org/pub/scm/linux/kernel/git/rusty/linux-2.6-cpumask: (36 commits)
cpumask: remove cpumask allocation from idle_balance, fix
numa, cpumask: move numa_node_id default implementation to topology.h, fix
cpumask: remove cpumask allocation from idle_balance
x86: cpumask: x86 mmio-mod.c use cpumask_var_t for downed_cpus
x86: cpumask: update 32-bit APM not to mug current->cpus_allowed
x86: microcode: cleanup
x86: cpumask: use work_on_cpu in arch/x86/kernel/microcode_core.c
cpumask: fix CONFIG_CPUMASK_OFFSTACK=y cpu hotunplug crash
numa, cpumask: move numa_node_id default implementation to topology.h
cpumask: convert node_to_cpumask_map[] to cpumask_var_t
cpumask: remove x86 cpumask_t uses.
cpumask: use cpumask_var_t in uv_flush_tlb_others.
cpumask: remove cpumask_t assignment from vector_allocation_domain()
cpumask: make Xen use the new operators.
cpumask: clean up summit's send_IPI functions
cpumask: use new cpumask functions throughout x86
x86: unify cpu_callin_mask/cpu_callout_mask/cpu_initialized_mask/cpu_sibling_setup_mask
cpumask: convert struct cpuinfo_x86's llc_shared_map to cpumask_var_t
cpumask: convert node_to_cpumask_map[] to cpumask_var_t
x86: unify 32 and 64-bit node_to_cpumask_map
...
04 Apr, 2009
1 commit
-
* 'ipi-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
s390: remove arch specific smp_send_stop()
panic: clean up kernel/panic.c
panic, smp: provide smp_send_stop() wrapper on UP too
panic: decrease oops_in_progress only after having done the panic
generic-ipi: eliminate WARN_ON()s during oops/panic
generic-ipi: cleanups
generic-ipi: remove CSD_FLAG_WAIT
generic-ipi: remove kmalloc()
generic IPI: simplify barriers and locking
03 Apr, 2009
1 commit
-
Impact: output all of packet commands - not just the first 4 / 8 bytes
Since commit d7e3c3249ef23b4617393c69fe464765b4ff1645 ("block: add
large command support"), struct request->cmd has been changed from
unsinged char cmd[BLK_MAX_CDB] to unsigned char *cmd.v1 -> v2: by: FUJITA Tomonori
- make sure rq->cmd_len is always intialized, and then we can use
rq->cmd_len instead of BLK_MAX_CDB.Signed-off-by: Li Zefan
Acked-by: FUJITA Tomonori
Cc: Arnaldo Carvalho de Melo
Cc: Steven Rostedt
Cc: Frederic Weisbecker
Cc: Jens Axboe
LKML-Reference:
Signed-off-by: Ingo Molnar
02 Apr, 2009
1 commit
-
Conflicts:
include/linux/slub_def.h
lib/Kconfig.debug
mm/slob.c
mm/slub.c
31 Mar, 2009
1 commit
-
Conflicts:
arch/x86/kernel/cpu/common.c
29 Mar, 2009
2 commits
-
…nux/kernel/git/tip/linux-2.6-tip
* 'percpu-cpumask-x86-for-linus-2' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip: (682 commits)
percpu: fix spurious alignment WARN in legacy SMP percpu allocator
percpu: generalize embedding first chunk setup helper
percpu: more flexibility for @dyn_size of pcpu_setup_first_chunk()
percpu: make x86 addr <-> pcpu ptr conversion macros generic
linker script: define __per_cpu_load on all SMP capable archs
x86: UV: remove uv_flush_tlb_others() WARN_ON
percpu: finer grained locking to break deadlock and allow atomic free
percpu: move fully free chunk reclamation into a work
percpu: move chunk area map extension out of area allocation
percpu: replace pcpu_realloc() with pcpu_mem_alloc() and pcpu_mem_free()
x86, percpu: setup reserved percpu area for x86_64
percpu, module: implement reserved allocation and use it for module percpu variables
percpu: add an indirection ptr for chunk page map access
x86: make embedding percpu allocator return excessive free space
percpu: use negative for auto for pcpu_setup_first_chunk() arguments
percpu: improve first chunk initial area map handling
percpu: cosmetic renames in pcpu_setup_first_chunk()
percpu: clean up percpu constants
x86: un-__init fill_pud/pmd/pte
x86: remove vestigial fix_ioremap prototypes
...Manually merge conflicts in arch/ia64/kernel/irq_ia64.c
-
* git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi-misc-2.6: (119 commits)
[SCSI] scsi_dh_rdac: Retry for NOT_READY check condition
[SCSI] mpt2sas: make global symbols unique
[SCSI] sd: Make revalidate less chatty
[SCSI] sd: Try READ CAPACITY 16 first for SBC-2 devices
[SCSI] sd: Refactor sd_read_capacity()
[SCSI] mpt2sas v00.100.11.15
[SCSI] mpt2sas: add MPT2SAS_MINOR(221) to miscdevice.h
[SCSI] ch: Add scsi type modalias
[SCSI] 3w-9xxx: add power management support
[SCSI] bsg: add linux/types.h include to bsg.h
[SCSI] cxgb3i: fix function descriptions
[SCSI] libiscsi: fix possbile null ptr session command cleanup
[SCSI] iscsi class: remove host no argument from session creation callout
[SCSI] libiscsi: pass session failure a session struct
[SCSI] iscsi lib: remove qdepth param from iscsi host allocation
[SCSI] iscsi lib: have lib create work queue for transmitting IO
[SCSI] iscsi class: fix lock dep warning on logout
[SCSI] libiscsi: don't cap queue depth in iscsi modules
[SCSI] iscsi_tcp: replace scsi_debug/tcp_debug logging with iscsi conn logging
[SCSI] libiscsi_tcp: replace tcp_debug/scsi_debug logging with session/conn logging
...
28 Mar, 2009
1 commit
-
Conflicts:
arch/parisc/kernel/irq.c
arch/x86/include/asm/fixmap_64.h
arch/x86/include/asm/setup.h
kernel/irq/handle.cSemantic merge:
arch/x86/include/asm/fixmap.hSigned-off-by: Ingo Molnar
26 Mar, 2009
2 commits
-
bsg submits REQ_TYPE_BLOCK_PC so the right check is max_hw_sectors.
But I've removed this check because right after, bsg proceeds with
calling blk_rq_map_user() which does all the right checks.Signed-off-by: Boaz Harrosh
Signed-off-by: Jens Axboe -
Put a WARN_ON in __blk_put_request if it is about to
leak bio(s). This is a serious bug that can happen in error
handling code paths.For this to work I have fixed a couple of places in block/ where
request->bio != NULL ownership was not honored. And a small cleanup
at sg_io() while at it.Signed-off-by: Boaz Harrosh
Signed-off-by: Jens Axboe
24 Mar, 2009
3 commits
-
Currently inherited from sg.c bsg will submit asynchronous request
at the head-of-the-queue, (using "at_head" set in the call to
blk_execute_rq_nowait()). This is bad in situation where the queues
are full, requests will execute out of order, and can cause
starvation of the first submitted requests.The sg_io_v4->flags member is used and a bit is allocated to denote the
Q_AT_TAIL. Zero is to queue at_head as before, to be compatible with old
code at the write/read path. SG_IO code path behavior was changed so to
be the same as write/read behavior. SG_IO was very rarely used and breaking
compatibility with it is OK at this stage.sg_io_hdr at sg.h also has a flags member and uses 3 bits from the first
nibble and one bit from the last nibble. Even though none of these bits
are supported by bsg, The second nibble is allocated for use by bsg. Just
in case.Signed-off-by: Boaz Harrosh
CC: Douglas Gilbert
Signed-off-by: Jens Axboe -
Signed-off-by: Jens Axboe
-
It calls blk_queue_make_request(), which sets the identical set of limits.
Signed-off-by: Jens Axboe
13 Mar, 2009
2 commits
-
Impact: cleanup
This is presumably what those definitions are for, and while all archs
define cpu_core_map/cpu_sibling map, that's changing (eg. x86 wants to
change it to a pointer).Signed-off-by: Rusty Russell
