Roland McGrath [Tue, 22 Apr 2008 19:21:25 +0000 (12:21 -0700)]
x86_64 ia32 ptrace: convert to compat_arch_ptrace
Now that there are no more special cases in sys32_ptrace, we
can convert to using the generic compat_sys_ptrace entry point.
The sys32_ptrace function gets simpler and becomes compat_arch_ptrace.
Signed-off-by: Roland McGrath <roland@redhat.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
Roland McGrath [Tue, 22 Apr 2008 19:20:20 +0000 (12:20 -0700)]
x86_64 ia32 ptrace: use compat_ptrace_request for siginfo
This removes the special-case handling for PTRACE_GETSIGINFO
and PTRACE_SETSIGINFO from x86_64's sys32_ptrace. The generic
compat_ptrace_request code handles these.
Signed-off-by: Roland McGrath <roland@redhat.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
Roland McGrath [Sat, 19 Apr 2008 21:27:56 +0000 (14:27 -0700)]
x86 signals: lift set_fs
This lifts the set_fs(USER_DS) call for signal handler setup out of the
three places copying the same code into the one place that calls them
all. There is no change in what it does.
Signed-off-by: Roland McGrath <roland@redhat.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
Roland McGrath [Sat, 19 Apr 2008 21:26:54 +0000 (14:26 -0700)]
x86 signals: lift flags diddling code
This lifts the code diddling the TF and DF bits for signal handler setup
out of the several places copying the same code into the one place that
calls them all. There is no change in what it does.
I also separated the recently-added DF bit clearing from the TF diddling.
The compiler turns them back into one instruction anyway. The tossing
in of DF to the same line of code with no new comments was a bit more
arcane than seems wise.
Signed-off-by: Roland McGrath <roland@redhat.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
Dmitri Vorobiev [Sun, 20 Apr 2008 20:47:55 +0000 (00:47 +0400)]
x86: remove NexGen support
It is claimed that NexGen CPUs were never shipped:
http://lkml.org/lkml/2008/4/20/179
Also, the kernel support for these chips has been broken for
a long time, the code intended to support NexGen thereby being
essentially dead.
As an outcome of the discussion that can be found using the URL
above, this patch removes the NexGen support altogether.
The changes in this patch survived a defconfig build for i386, a
couple of successful randconfig builds, as well as a runtime test,
which consisted in booting a 32-bit x86 box up to the shell prompt.
Signed-off-by: Dmitri Vorobiev <dmitri.vorobiev@gmail.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
Dmitri Vorobiev [Sun, 20 Apr 2008 02:54:33 +0000 (06:54 +0400)]
x86: array can become static
In arch/x86/kernel/setup_64.c, the standard_io_resources array
is needlessly defined as global. This patch makes this variable
static.
This patch was successfully build-tested using the defconfig
for x86_64. Runtime test was performed by booting a 64-bit x86
box up to the shell prompt.
Signed-off-by: Dmitri Vorobiev <dmitri.vorobiev@gmail.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
Dmitri Vorobiev [Sun, 20 Apr 2008 02:54:31 +0000 (06:54 +0400)]
x86: remove unused function amd_init_cpu()
There are no users for the function amd_init_cpu() defined in
arch/x86/kernel/cpu/amd.c. This patch removes this routine.
This patch was build-tested using defconfigs for i386 and x86_64,
and a few randconfig instances. Runtime tests were performed by
booting 32- and 64-bit x86 boxen up to the shell prompt.
Signed-off-by: Dmitri Vorobiev <dmitri.vorobiev@gmail.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
Jan Beulich [Tue, 22 Apr 2008 15:22:21 +0000 (16:22 +0100)]
x86-64: extend MCE CPU quirk handling
At least on my Barcelona, I see MCE log entries after cold boot caused
by BIOS not properly clearing the respective registers. Therefore, this
patch extends the workaround to families 0x10 and 0x11 (the latter just
for completeness, I have nothing to verify this against).
At the same time, provide a way to make these entries visible via the
'mce=bootlog' command line option even on these machines.
Signed-off-by: Jan Beulich <jbeulich@novell.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
Jan Beulich [Tue, 22 Apr 2008 15:28:41 +0000 (16:28 +0100)]
x86: fix watchdog ops for CoreDuo
There apparently was an unnoticed conflict between an earlier patch to
this file and mine (d1e084746b0e5806e6345ab31c5b370f8dee2b23), which
I noticed only now. I suppose a change like the one below (untested) is
needed; I didn't get any response on a confirmation request for this from
the submitter of the first patch.
The issue is the writing of the 'checkbit' member at the end of
setup_intel_arch_watchdog(), which my patch made go to intel_arch_wd_ops
rather than wd_ops.
Signed-off-by: Jan Beulich <jbeulich@novell.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
Roland Dreier [Fri, 25 Apr 2008 22:32:51 +0000 (15:32 -0700)]
Export __locks_copy_lock() so modular lockd builds
Commit 1a747ee0 ("locks: don't call ->copy_lock methods on return of
conflicting locks") changed fs/lockd/svclock.c to call
__locks_copy_lock() instead of locks_copy_lock(), but lockd can be built
as a module and __locks_copy_lock() is not exported, which causes a
build error
Merge branch 'master' of git://git.kernel.org/pub/scm/linux/kernel/git/paulus/powerpc
* 'master' of git://git.kernel.org/pub/scm/linux/kernel/git/paulus/powerpc: (49 commits)
[POWERPC] Add zImage.iseries to arch/powerpc/boot/.gitignore
[POWERPC] bootwrapper: fix build error on virtex405-head.S
[POWERPC] 4xx: Fix 460GT support to not enable FPU
[POWERPC] 4xx: Add NOR FLASH entries to Canyonlands and Glacier dts
[POWERPC] Xilinx: of_serial support for Xilinx uart 16550.
[POWERPC] Xilinx: boot support for Xilinx uart 16550.
[POWERPC] celleb: Add support for PCI Express
[POWERPC] celleb: Move miscellaneous files for Beat
[POWERPC] celleb: Move a file for SPU on Beat
[POWERPC] celleb: Move files for Beat mmu and iommu
[POWERPC] celleb: Move files for Beat hvcall interfaces
[POWERPC] celleb: Move the SCC related code for celleb
[POWERPC] celleb: Move the files for celleb base support
[POWERPC] celleb: Consolidate io-workarounds code
[POWERPC] cell: Generalize io-workarounds code
[POWERPC] Add CONFIG_PPC_PSERIES_DEBUG to enable debugging for platforms/pseries
[POWERPC] Convert from DBG() to pr_debug() in platforms/pseries/
[POWERPC] Register udbg console early on pseries LPAR
[POWERPC] Mark udbg console as CON_ANYTIME, ie. callable early in boot
[POWERPC] Set udbg_console index to 0
...
Merge branch 'for-linus' of git://git.o-hand.com/linux-rpurdie-leds
* 'for-linus' of git://git.o-hand.com/linux-rpurdie-leds:
leds: Add default-on trigger
leds: Document the context brightness_set needs
leds: Add new driver for the LEDs on the Freecom FSG-3
leds: Add support to leds with readable status
leds: enable support for blink_set() platform hook in leds-gpio
leds: Cleanup various whitespace and code style issues
leds: disable triggers on brightness set
leds: Add mail LED support for "Clevo D400P"
Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/x86/linux-2.6-x86-pat
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/x86/linux-2.6-x86-pat:
generic: add ioremap_wc() interface wrapper
/dev/mem: make promisc the default
pat: cleanups
x86: PAT use reserve free memtype in mmap of /dev/mem
x86: PAT phys_mem_access_prot_allowed for dev/mem mmap
x86: PAT avoid aliasing in /dev/mem read/write
devmem: add range_is_allowed() check to mmap of /dev/mem
x86: introduce /dev/mem restrictions with a config option
Merge branch 'upstream-linus' of 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:
pata_bf54x: decrease count first.
sata_mv: re-enable hotplug, update TODO list
sata_mv: leave SError bits untouched in mv_err_intr
sata_mv: more interrupt handling rework
sata_mv: tidy host controller interrupt handling
sata_mv: simplify request/response queue handling
sata_mv: simplify freeze/thaw bit-shift calculations
sata_mv mask all interrupt coalescing bits
sata_mv more cosmetics
ata_piix: add Asus Eee 701 controller to short cable list
libata-eh set tf flags in NCQ EH result_tf
make sata_set_spd_needed() static
make sata_print_link_status() static
libata-acpi.c: remove unneeded #if's
sata_nv: make hardreset return -EAGAIN on success
ahci: retry enabling AHCI a few times before spitting out WARN_ON()
libata: make WARN_ON conditions in ata_sff_hsm_move() more strict
ATA/IDE: fix platform driver hotplug/coldplug
sata_sis: SCR accessors return -EINVAL when requested SCR isn't available
libata: functions with definition should not be extern
* git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb-2.6: (120 commits)
usb: don't update devnum for wusb devices
wusb: make ep0_reinit available for modules
wusb: devices dont use a set address
wusb: teach choose_address() about wireless devices
wusb: add link wusb-usb device
wusb: add authenticathed bit to usb_dev
USB: remove unnecessary type casting of urb->context
usb serial: more fixes and groundwork for tty changes
USB: replace remaining __FUNCTION__ occurrences
USB: usbfs: export the URB_NO_INTERRUPT flag to userspace
USB: fix compile problems in ehci-hcd
USB: ehci: qh_completions cleanup and bugfix
USB: cdc-acm: signedness fix
USB: add documentation about callbacks
USB: don't explicitly reenable root-hub status interrupts
USB: OHCI: turn off RD when remote wakeup is disabled
USB: HCDs use the do_remote_wakeup flag
USB: g_file_storage: ignore bulk-out data after invalid CBW
USB: serial: remove endpoints setting checks from core and header
USB: serial: remove unneeded number endpoints settings
...
* git://git.kernel.org/pub/scm/linux/kernel/git/sam/kbuild:
kconfig: fix broken target update-po-config
kbuild: silence documentation GEN xml messages according to $(quiet)
Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/x86/linux-2.6-xen-next
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/x86/linux-2.6-xen-next: (52 commits)
xen: add balloon driver
xen: allow compilation with non-flat memory
xen: fold xen_sysexit into xen_iret
xen: allow set_pte_at on init_mm to be lockless
xen: disable preemption during tlb flush
xen pvfb: Para-virtual framebuffer, keyboard and pointer driver
xen: Add compatibility aliases for frontend drivers
xen: Module autoprobing support for frontend drivers
xen blkfront: Delay wait for block devices until after the disk is added
xen/blkfront: use bdget_disk
xen: Make xen-blkfront write its protocol ABI to xenstore
xen: import arch generic part of xencomm
xen: make grant table arch portable
xen: replace callers of alloc_vm_area()/free_vm_area() with xen_ prefixed one
xen: make include/xen/page.h portable moving those definitions under asm dir
xen: add resend_irq_on_evtchn() definition into events.c
Xen: make events.c portable for ia64/xen support
xen: move events.c to drivers/xen for IA64/Xen support
xen: move features.c from arch/x86/xen/features.c to drivers/xen
xen: add missing definitions in include/xen/interface/vcpu.h which ia64/xen needs
...
* git://git.infradead.org/mtd-2.6: (82 commits)
[MTD] m25p80: Add Support for ATMEL AT25DF641 64-Megabit SPI Flash
[MTD] m25p80: add FAST_READ access support to M25Pxx
[MTD] [NAND] bf5xx_nand: Avoid crash if bfin_mac is installed.
[MTD] [NAND] at91_nand: control NCE signal
[MTD] [NAND] AT91 hardware ECC compile fix for at91sam9263 / at91sam9260
[MTD] [NAND] Hardware ECC controller on at91sam9263 / at91sam9260
[JFFS2] Introduce dbg_readinode2 log level, use it to shut read_dnode() up
[JFFS2] Fix jffs2_reserve_space() when all blocks are pending erasure.
[JFFS2] Add erase_checking_list to hold blocks being marked.
UBI: add a message
[JFFS2] Return values of jffs2_block_check_erase error paths
[MTD] Clean up AR7 partition map support
[MTD] [NOR] Fix Intel CFI driver for collie flash
[JFFS2] Finally remove redundant ref->__totlen field.
[JFFS2] Honour TEST_TOTLEN macro in debugging code. ref->__totlen is going!
[JFFS2] Add paranoia debugging for superblock counts
[JFFS2] Fix free space leak with in-band cleanmarkers
[JFFS2] Self-sufficient #includes in jffs2_fs_i.h: include <linux/mutex.h>
[MTD] [NAND] Verify probe by retrying to checking the results match
[MTD] [NAND] S3C2410 Allow ECC disable to be specified by the board
...
Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/ieee1394/linux1394-2.6
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/ieee1394/linux1394-2.6:
ieee1394: silence defined but not used warning in non-modular builds
ieee1394: rawiso: requeue packet for transmission after skipped cycle
Sam Ravnborg [Fri, 25 Apr 2008 19:15:41 +0000 (21:15 +0200)]
kconfig: fix broken target update-po-config
Massimo Maiurana reported:
In the latest kernel "make update-po-config" fails because it tries
to open arch/Kconfig/Kconfig, since the ls command doesn't
distinguish between files and directories.
Cc: Massimo Maiurana <maiurana@gmail.com> Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
* git://git.kernel.org/pub/scm/linux/kernel/git/sam/kbuild:
kbuild: fix depmod comment
kbuild: Add new Kbuild variable KBUILD_EXTRA_SYMBOLS
kbuild: support loading extra symbols in modpost
Add option to enable -Wframe-larger-than= on gcc 4.4
kbuild: add kconfig symbols to tags output
kbuild: fix some minor typoes
kbuild: error out on missing MODULE_LICENSE
Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mingo/linux-2.6-x86-fixes4
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mingo/linux-2.6-x86-fixes4:
x86: harden kernel code patching
x86: clean up text_poke()
x86: fix text_poke()
x86: remove set_fixmap() warning
x86: make __set_fixmap() non-init
x86: make clear_fixmap() available on 64-bit as well
Mike Frysinger [Fri, 28 Mar 2008 21:30:58 +0000 (14:30 -0700)]
kbuild: silence documentation GEN xml messages according to $(quiet)
Add rules for gen_xml and its quiet & silent variants. This causes "make -s"
to be silent for gen_xml.
Signed-off-by: Mike Frysinger <vapier@gentoo.org> Acked-by: Randy Dunlap <randy.dunlap@oracle.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
Merge branch 'for-linus' of git://linux-nfs.org/~bfields/linux
* 'for-linus' of git://linux-nfs.org/~bfields/linux:
nfsd: don't allow setting ctime over v4
Update to NFS/RDMA documentation
locks: don't call ->copy_lock methods on return of conflicting locks
lockd: unlock lockd locks held for a certain filesystem
lockd: unlock lockd locks associated with a given server ip
leases: remove unneeded variable from fcntl_setlease().
leases: move lock allocation earlier in generic_setlease()
leases: when unlocking, skip locking-related steps
leases: fix a return-value mixup
Richard Hacker [Thu, 28 Feb 2008 08:40:58 +0000 (09:40 +0100)]
kbuild: Add new Kbuild variable KBUILD_EXTRA_SYMBOLS
This patch adds a new (Kbuild) Makefile variable KBUILD_EXTRA_SYMBOLS.
The space separated list of file names assigned to KBUILD_EXTRA_SYMBOLS
is used when calling scripts/mod/modpost during stage 2 of the Kbuild
process for non-kernel-tree modules.
Signed-off-by: Richard Hacker <lerichi@gmx.net> Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
Richard Hacker [Thu, 28 Feb 2008 08:40:52 +0000 (09:40 +0100)]
kbuild: support loading extra symbols in modpost
This patch adds a new command line option -E to modpost, expecting a symbol
file as an argument which is read prior to symbol processing. -E can be
supplied multiple times for as many files as is needed.
When building kernel modules that depend on other modules not in the main
kernel tree, modpost complains about undefined symbols:
# make -C /path/to/linux/kernel M=/path/to/my/module
...
Building modules, stage 2.
....
WARNING: "rt_copy_buf" [/home/rich/osc_etl_rtw/osc_kmod.ko] undefined!
...etc
This situation occurs when modpost processes the new module's symbols. When
it finds symbols not exported by the mainline kernel, it issues this warning.
The patch adds a new command line option -e to modpost which expects a symbol
file as an argument. The symbols listed in this file are added to modpost's
symbol tables during startup. -e can be supplied as often as required.
This patch works together with the second patch. It introduces a new make
variable, KBUILD_EXTRA_SYMBOLS, which is used when calling modpost.
Signed-off-by: Richard Hacker <lerichi@gmx.net> Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
Andi Kleen [Fri, 22 Feb 2008 14:15:03 +0000 (15:15 +0100)]
Add option to enable -Wframe-larger-than= on gcc 4.4
Add option to enable -Wframe-larger-than= on gcc 4.4
gcc mainline (upcoming 4.4) added a new -Wframe-larger-than=...
option to warn at build time about too large stack frames. Add a config
option to enable this warning, since this very useful for the kernel.
I choose (somewhat arbitarily) 2048 as default warning threshold for 64bit
and 1024 as default for 32bit architectures. With some research and
fixing all the code for smaller values these defaults should be probably
lowered.
With the default allyesconfigs have some new warnings, but I think
that is all code that should be just fixed.
At some point (when gcc 4.4 is released and widely used) this should
obsolete make checkstack
Signed-off-by: Andi Kleen <ak@suse.de> Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
Sam Ravnborg [Fri, 25 Apr 2008 18:13:30 +0000 (20:13 +0200)]
kbuild: error out on missing MODULE_LICENSE
Adrian Bunk suggested a build time check for
missing MODULE_LICENSE annotation in modules.
The build time check is fatal as we really
want this fixed for all modules.
In-tree modules should all have been fixed up by now.
Signed-off-by: Sam Ravnborg <sam@ravnborg.org> Cc: Adrian Bunk <bunk@kernel.org>
J. Bruce Fields [Thu, 24 Apr 2008 20:59:30 +0000 (16:59 -0400)]
nfsd: don't allow setting ctime over v4
Presumably this is left over from earlier drafts of v4, which listed
TIME_METADATA as writeable. It's read-only in rfc 3530, and shouldn't
be modifiable anyway.
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
J. Bruce Fields [Thu, 24 Apr 2008 14:08:22 +0000 (10:08 -0400)]
locks: don't call ->copy_lock methods on return of conflicting locks
The file_lock structure is used both as a heavy-weight representation of
an active lock, with pointers to reference-counted structures, etc., and
as a simple container for parameters that describe a file lock.
The conflicting lock returned from __posix_lock_file is an example of
the latter; so don't call the filesystem or lock manager callbacks when
copying to it. This also saves the need for an unnecessary
locks_init_lock in the nfsv4 server.
Thanks to Trond for pointing out the error.
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu> Cc: Trond Myklebust <Trond.Myklebust@netapp.com>
Wendy Cheng [Thu, 17 Jan 2008 16:10:12 +0000 (11:10 -0500)]
lockd: unlock lockd locks associated with a given server ip
For high-availability NFS service, we generally need to be able to drop
file locks held on the exported filesystem before moving clients to a
new server. Currently the only way to do that is by shutting down lockd
entirely, which is often undesireable (for example, if you want to
continue exporting other filesystems).
This patch allows the administrator to release all locks held by clients
accessing the client through a given server ip address, by echoing that
address to a new file, /proc/fs/nfsd/unlock_ip, as in:
shell> echo 10.1.1.2 > /proc/fs/nfsd/unlock_ip
The expected sequence of events can be:
1. Tear down the IP address
2. Unexport the path
3. Write IP to /proc/fs/nfsd/unlock_ip to unlock files
4. Signal peer to begin take-over.
For now we only support IPv4 addresses and NFSv2/v3 (NFSv4 locks are not
affected).
Also, if unmounting the filesystem is required, we assume at step 3 that
clients using the given server ip are the only clients holding locks on
the given filesystem; otherwise, an additional patch is required to
allow revoking all locks held by lockd on a given filesystem.
Signed-off-by: S. Wendy Cheng <wcheng@redhat.com> Cc: Lon Hohberger <lhh@redhat.com> Cc: Christoph Hellwig <hch@lst.de> Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
fs/lockd/svcsubs.c | 66 +++++++++++++++++++++++++++++++++++++++-----
fs/nfsd/nfsctl.c | 65 +++++++++++++++++++++++++++++++++++++++++++
include/linux/lockd/lockd.h | 7 ++++
3 files changed, 131 insertions(+), 7 deletions(-)
David M. Richter [Wed, 23 Apr 2008 20:29:01 +0000 (16:29 -0400)]
leases: move lock allocation earlier in generic_setlease()
In generic_setlease(), the struct file_lock is allocated after tests for the
presence of conflicting readers/writers is done, despite the fact that the
allocation might block; this patch moves the allocation earlier. A subsequent
set of patches will rely on this behavior to properly serialize between a
modified __break_lease() and generic_setlease().
Signed-off-by: David M. Richter <richterd@citi.umich.edu> Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
David M. Richter [Wed, 23 Apr 2008 20:28:59 +0000 (16:28 -0400)]
leases: fix a return-value mixup
Fixes a return-value mixup from 85c59580b30c82aa771aa33b37217a6b6851bc14
"locks: Fix potential OOPS in generic_setlease()", in which -ENOMEM replaced
what had been intended to stay -EAGAIN in the variable "error".
Signed-off-by: David M. Richter <richterd@citi.umich.edu> Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
Tony Breeds [Thu, 24 Apr 2008 07:02:04 +0000 (09:02 +0200)]
ieee1394: silence defined but not used warning in non-modular builds
Currently the kernel will issue the following warning:
drivers/ieee1394/raw1394.c:2938: warning: 'raw1394_id_table' defined but not used
Add #ifdef MODULE guards around the declaration.
Signed-off-by: Tony Breeds <tony@bakeyournoodle.com>
Ditto with dv1394_id_table and video1394_id_table.
Signed-off-by: Stefan Richter <stefanr@s5r6.in-berlin.de>
Pieter Palmers [Wed, 19 Mar 2008 21:10:59 +0000 (22:10 +0100)]
ieee1394: rawiso: requeue packet for transmission after skipped cycle
As it seems, some host controllers have issues that can cause them to
skip cycles now and then when using large packets. I suspect that this
is due to DMA not succeeding in time. If the transmit fifo can't contain
more than one packet (big packets), the DMA should provide a new packet
each cycle (125us). I am under the impression that my current PCI
express test system can't guarantee this.
In any case, the patch tries to provide a workaround as follows:
The DMA program descriptors are modified such that when an error occurs,
the DMA engine retries the descriptor the next cycle instead of
stalling. This way no data is lost. The side effect of this is that
packets are sent with one cycle delay. This however might not be that
much of a problem for certain protocols (e.g. AM824). If they use
padding packets for e.g. rate matching they can drop one of those to
resync the streams.
The amount of skips between two userspace wakeups is counted. This
number is then propagated to userspace through the upper 16 bits of the
'dropped' parameter. This allows unmodified userspace applications due
to the following:
1) libraw simply passes this dropped parameter to the user application
2) the meaning of the dropped parameter is: if it's nonzero, something
bad has happened. The actual value of the parameter at this moment does
not have a specific meaning.
A libraw client can then retrieve the number of skipped cycles and
account for them if needed.
Signed-off-by: Pieter Palmers <pieterp@joow.be> Signed-off-by: Stefan Richter <stefanr@s5r6.in-berlin.de>
Al Viro [Thu, 24 Apr 2008 00:38:10 +0000 (20:38 -0400)]
[PATCH] sanitize locate_fd()
* 'file' argument is unused; lose it.
* move setting flags from the caller (dupfd()) to locate_fd();
pass cloexec flag as new argument. Note that files_fdtable()
that used to be in dupfd() isn't needed in the place in
locate_fd() where the moved code ends up - we know that ->file_lock
hadn't been dropped since the last time we calculated fdt because
we can get there only if expand_files() returns 0 and it doesn't
drop/reacquire in that case.
* move getting/dropping ->file_lock into locate_fd(). Now the caller
doesn't need to do anything with files_struct *files anymore and
we can move that inside locate_fd() as well, killing the
struct files_struct * argument.
At that point locate_fd() is extremely similar to get_unused_fd_flags()
and the next patches will merge those two.
Al Viro [Tue, 22 Apr 2008 09:31:30 +0000 (05:31 -0400)]
[PATCH] sanitize unshare_files/reset_files_struct
* let unshare_files() give caller the displaced files_struct
* don't bother with grabbing reference only to drop it in the
caller if it hadn't been shared in the first place
* in that form unshare_files() is trivially implemented via
unshare_fd(), so we eliminate the duplicate logics in fork.c
* reset_files_struct() is not just only called for current;
it will break the system if somebody ever calls it for anything
else (we can't modify ->files of somebody else). Lose the
task_struct * argument.
Al Viro [Tue, 22 Apr 2008 09:11:59 +0000 (05:11 -0400)]
[PATCH] sanitize handling of shared descriptor tables in failing execve()
* unshare_files() can fail; doing it after irreversible actions is wrong
and de_thread() is certainly irreversible.
* since we do it unconditionally anyway, we might as well do it in do_execve()
and save ourselves the PITA in binfmt handlers, etc.
* while we are at it, binfmt_som actually leaked files_struct on failure.
As a side benefit, unshare_files(), put_files_struct() and reset_files_struct()
become unexported.
The small patch below:
- Removes the unused md argument from both specific_minor() and next_free_minor()
- Folds kmalloc + memset(0) into a single kzalloc call in alloc_dev()
This has been compile tested on x86.
Signed-off-by: Frederik Deweerdt <frederik.deweerdt@gmail.com> Signed-off-by: Alasdair G Kergon <agk@redhat.com>
Remove an avoidable 3ms delay on some dm-raid1 and kcopyd I/O.
It is specified that any submitted bio without BIO_RW_SYNC flag may plug the
queue (i.e. block the requests from being dispatched to the physical device).
The queue is unplugged when the caller calls blk_unplug() function. Usually, the
sequence is that someone calls submit_bh to submit IO on a buffer. The IO plugs
the queue and waits (to be possibly joined with other adjacent bios). Then, when
the caller calls wait_on_buffer(), it unplugs the queue and submits the IOs to
the disk.
This was happenning:
When doing O_SYNC writes, function fsync_buffers_list() submits a list of
bios to dm_raid1, the bios are added to dm_raid1 write queue and kmirrord is
woken up.
fsync_buffers_list() calls wait_on_buffer(). That unplugs the queue, but
there are no bios on the device queue as they are still in the dm_raid1 queue.
wait_on_buffer() starts waiting until the IO is finished.
kmirrord is scheduled, kmirrord takes bios and submits them to the devices.
The submitted bio plugs the harddisk queue but there is no one to unplug it.
(The process that called wait_on_buffer() is already sleeping.)
So there is a 3ms timeout, after which the queues on the harddisks are
unplugged and requests are processed.
This 3ms timeout meant that in certain workloads (e.g. O_SYNC, 8kb writes),
dm-raid1 is 10 times slower than md raid1.
Every time we submit something asynchronously via dm_io, we must unplug the
queue actually to send the request to the device.
This patch adds an unplug call to kmirrord - while processing requests, it keeps
the queue plugged (so that adjacent bios can be merged); when it finishes
processing all the bios, it unplugs the queue to submit the bios.
It also fixes kcopyd which has the same potential problem. All kcopyd requests
are submitted with BIO_RW_SYNC.
This patch replaces the schedule() in the main kmirrord thread with a timer.
The schedule() could introduce an unwanted delay when work is ready to be
processed.
The code instead calls wake() when there's work to be done immediately, and
delayed_wake() after a failure to give a short delay before retrying.
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Signed-off-by: Alasdair G Kergon <agk@redhat.com>
Remove client counting code that is no longer needed.
Initialization and destruction is made globally from dm_init and dm_exit and is
not based on client counts. Initialization allocates only one empty slab cache,
so there is no negative impact from performing the initialization always,
regardless of whether some client uses kcopyd or not.
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Signed-off-by: Alasdair G Kergon <agk@redhat.com>
Milan Broz [Thu, 24 Apr 2008 20:42:36 +0000 (21:42 +0100)]
dm snapshot: reduce default memory allocation
Limit the amount of memory allocated per snapshot on systems
with a large page size. (The larger default chunk size on
these systems compensates for the smaller number of pages reserved.)
Signed-off-by: Milan Broz <mbroz@redhat.com> Signed-off-by: Alasdair G Kergon <agk@redhat.com>
If a snapshot has a smaller chunksize than the page size the
conversion to pages currently returns 0 instead of 1, causing:
kernel BUG in mempool_resize.
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Signed-off-by: Milan Broz <mbroz@redhat.com> Signed-off-by: Alasdair G Kergon <agk@redhat.com> Cc: stable@kernel.org
projects / linux-2.6 / log