Izik Eidus [Sun, 11 Nov 2007 12:40:48 +0000 (14:40 +0200)]
KVM: x86 emulator: fix JMP_REL
Change JMP_REL to call to register_address_increment(): the operands size
should not effect the calculation of the eip, instead the ad_bytes should
affect it.
Signed-off-by: Izik Eidus <izike@qumranet.com> Signed-off-by: Avi Kivity <avi@qumranet.com>
* git://git.kernel.org/pub/scm/linux/kernel/git/herbert/net-2.6: (41 commits)
[XFRM]: Fix leak of expired xfrm_states
[ATM]: [he] initialize lock and tasklet earlier
[IPV4]: Remove bogus ifdef mess in arp_process
[SKBUFF]: Free old skb properly in skb_morph
[IPV4]: Fix memory leak in inet_hashtables.h when NUMA is on
[IPSEC]: Temporarily remove locks around copying of non-atomic fields
[TCP] MTUprobe: Cleanup send queue check (no need to loop)
[TCP]: MTUprobe: receiver window & data available checks fixed
[MAINTAINERS]: tlan list is subscribers-only
[SUNRPC]: Remove SPIN_LOCK_UNLOCKED
[SUNRPC]: Make xprtsock.c:xs_setup_{udp,tcp}() static
[PFKEY]: Sending an SADB_GET responds with an SADB_GET
[IRDA]: Compilation for CONFIG_INET=n case
[IPVS]: Fix compiler warning about unused register_ip_vs_protocol
[ARP]: Fix arp reply when sender ip 0
[IPV6] TCPMD5: Fix deleting key operation.
[IPV6] TCPMD5: Check return value of tcp_alloc_md5sig_pool().
[IPV4] TCPMD5: Use memmove() instead of memcpy() because we have overlaps.
[IPV4] TCPMD5: Omit redundant NULL check for kfree() argument.
ieee80211: Stop net_ratelimit/IEEE80211_DEBUG_DROP log pollution
...
* git://git.kernel.org/pub/scm/linux/kernel/git/herbert/crypto-2.6:
[CRYPTO] api: Fix potential race in crypto_remove_spawn
[CRYPTO] authenc: Move initialisations up to shut up gcc
Linus Torvalds [Tue, 27 Nov 2007 03:42:59 +0000 (19:42 -0800)]
Merge git://git.linux-nfs.org/pub/linux/nfs-2.6
* git://git.linux-nfs.org/pub/linux/nfs-2.6:
NFS: Clean up new multi-segment direct I/O changes
NFS: Ensure we return zero if applications attempt to write zero bytes
NFS: Support multiple segment iovecs in the NFS direct I/O path
NFS: Introduce iovec I/O helpers to fs/nfs/direct.c
SUNRPC: Add missing "space" to net/sunrpc/auth_gss.c
SUNRPC: make sunrpc/xprtsock.c:xs_setup_{udp,tcp}() static
NFS: fs/nfs/dir.c should #include "internal.h"
NFS: make nfs_wb_page_priority() static
NFS: mount failure causes bad page state
SUNRPC: remove NFS/RDMA client's binary sysctls
kernel BUG at fs/nfs/namespace.c:108! - can be triggered by bad server
sunrpc: rpc_pipe_poll may miss available data in some cases
sunrpc: return error if unsupported enctype or cksumtype is encountered
sunrpc: gss_pipe_downcall(), don't assume all errors are transient
NFS: Fix the ustat() regression
* git://git.kernel.org/pub/scm/linux/kernel/git/x86/linux-2.6-x86:
x86: fix APIC related bootup crash on Athlon XP CPUs
time: add ADJ_OFFSET_SS_READ
x86: export the symbol empty_zero_page on the 32-bit x86 architecture
x86: fix kprobes_64.c inlining borkage
pci: use pci=bfsort for HP DL385 G2, DL585 G2
x86: correctly set UTS_MACHINE for "make ARCH=x86"
lockdep: annotate do_debug() trap handler
x86: turn off iommu merge by default
x86: fix ACPI compile for LOCAL_APIC=n
x86: printk kernel version in WARN_ON and other dump_stack users
ACPI: Set max_cstate to 1 for early Opterons.
x86: fix NMI watchdog & 'stopped time' problem
Linus Torvalds [Tue, 27 Nov 2007 03:40:27 +0000 (19:40 -0800)]
Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/cooloney/blackfin-2.6
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/cooloney/blackfin-2.6: (56 commits)
Blackfin arch: fix bug when enable uart1 with uart0 disabled => no initial console
Blackfin arch: split apart dump_bfin_regs and merge/remove show_regs from process.c, which was largely duplicated
Blackfin arch: use common __INIT/__FINIT defines rather than setting the .section ourselves to .init.text
Blackfin arch: fix bug when sending signals with the wrong PC, cause gdb get confused
Blackfin arch: Ensure we printk out strings with the proper loglevel
Blackfin arch: Need to specify ax with the .init.text section,
Blackfin arch: Update Kconfig to latest Blackfin silicon datasheets
Blackfin arch: update defconfig files
Blackfin arch: Fix typo, and add ENDPROC - no functional changes
Blackfin arch: convert READY to DMA_READY as it causes build errors in common sound code otherwise
Blackfin arch: add defines for the on-chip L1 ROM of BF54x
Blackfin arch: cplb and map header file cleanup
Blackfin arch: cleanup the cplb declares
Blackfin arch: fix broken on BF52x, remove silly checks on processors for L1_SCRATCH defines
Blackfin arch: add support for working around anomaly 05000312
Blackfin arch: cleanup BF54x header file and add BF547 definition
Blackfin arch: fix building for BF542 processors which only have 1 TWI
Blackfin arch: rename _return_from_exception to _bfin_return_from_exception and export it
Blackfin arch: move EXPORT_SYMBOL() to C files where the symbol is actually defined
Blackfin arch: fix bug NOR Flash MTD mount fail
...
* git://git.kernel.org/pub/scm/linux/kernel/git/lethal/sh-2.6.24:
sh: include ax88796 in the defconfig for r7785rp
sh: include ax88796 in the defconfig for r7780mp
sh: fix R2D-1 CF support
fb: Orphan imsttfb.
sh: Update mailing list info.
sh: lockless UTLB miss fast-path.
sh: Kill off UTLB flush in fast-path.
sh: Fix copy_{to,from}_user_page() with cache disabled.
Mikulas Patocka [Thu, 22 Nov 2007 20:26:01 +0000 (21:26 +0100)]
plip: fix parport_register_device name parameter
Plip passes a string "name" that is allocated on stack to
parport_register_device. parport_register_device holds the pointer to
"name" and when the registering function exits, it points nowhere.
On some machine, this bug causes bad names to appear in /proc, such as
/proc/sys/dev/parport/parport0/devices/T^/�X^/�, on others, the plip
proc node is completely missing.
The patch also fixes documentation to note this requirement.
Mikulas Patocka [Thu, 22 Nov 2007 20:19:39 +0000 (21:19 +0100)]
plip: use netif_rx_ni() for packet receive
netif_rx is meant to be called from interrupts because it doesn't wake
up ksoftirqd. For calling from outside interrupts, netif_rx_ni exists.
This fixes plip to use netif_rx_ni. It fixes the infamous error "NOHZ:
local_softirq_panding 08" that happens on some machines with NOHZ and
plip --- it is caused by the fact that softirq is pending and ksoftirqd
is sleeping.
Linus Torvalds [Tue, 27 Nov 2007 03:25:23 +0000 (19:25 -0800)]
Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/drzeus/mmc
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/drzeus/mmc:
mmc: Avoid re-using minor numbers before the original device is closed.
tifm_sd: handle non-power-of-2 block sizes
mmc_block: check card state after write
* git://git.kernel.org/pub/scm/linux/kernel/git/sfrench/cifs-2.6:
[CIFS] Fix check after use error in ACL code
[CIFS] Fix potential data corruption when writing out cached dirty pages
[CIFS] Fix spurious reconnect on 2nd peek from read of SMB length
[CIFS] remove build warning
[CIFS] Have CIFS_SessSetup build correct SPNEGO SessionSetup request
[CIFS] minor checkpatch cleanup
[CIFS] have cifs_get_spnego_key get the hostname from TCP_Server_Info
[CIFS] add hostname field to TCP_Server_Info struct
[CIFS] clean up error handling in cifs_mount
[CIFS] add ver= prefix to upcall format version
[CIFS] Fix buffer overflow if server sends corrupt response to small
Linus Torvalds [Tue, 27 Nov 2007 03:18:22 +0000 (19:18 -0800)]
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: (21 commits)
libata: bump transfer chunk size if it's odd
libata: Return proper ATA INT status in pata_bf54x driver
pata_ali: trim trailing whitespace (fix checkpatch complaints)
pata_isapnp: Polled devices
pata_hpt37x: Fix cable detect bug spotted by Sergei
pata_ali: Lots of problems still showing up with small ATAPI DMA
pata_ali: Add Mitac 8317 and derivatives
libata-core: List more documentation sources for reference
ata_piix: Invalid use of writel/readl with iomap
sata_sil24: fix sg table sizing
pata_jmicron: fix disabled port handling in jmicron_pre_reset()
pata_sil680: kill bogus reset code (take 2)
ata_piix: port enable for the first SATA controller of ICH8 is 0xf not 0x3
ata_piix: only enable the first port on apple macbook pro
ata_piix: reorganize controller IDs
pata_sis.c: Add Packard Bell EasyNote K5305 to laptops
libata-scsi: be tolerant of 12-byte ATAPI commands in 16-byte CDBs
libata: use ATA_HORKAGE_STUCK_ERR for ATAPI tape drives
libata: workaround DRQ=1 ERR=1 for ATAPI tape drives
libata: remove unused functions
...
* git://git.kernel.org/pub/scm/linux/kernel/git/rusty/linux-2.6-for-linus:
virtio: fix net driver loop case where we fail to restart
module: fix and elaborate comments
virtio: fix module/device unloading
lguest: Fix uninitialized members in example launcher
Patrick McHardy [Tue, 27 Nov 2007 03:10:07 +0000 (11:10 +0800)]
[XFRM]: Fix leak of expired xfrm_states
The xfrm_timer calls __xfrm_state_delete, which drops the final reference
manually without triggering destruction of the state. Change it to use
xfrm_state_put to add the state to the gc list when we're dropping the
last reference. The timer function may still continue to use the state
safely since the final destruction does a del_timer_sync().
Signed-off-by: Patrick McHardy <kaber@trash.net> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Linus Torvalds [Tue, 27 Nov 2007 03:09:58 +0000 (19:09 -0800)]
Merge branch 'release' of git://git.kernel.org/pub/scm/linux/kernel/git/lenb/linux-acpi-2.6
* 'release' of git://git.kernel.org/pub/scm/linux/kernel/git/lenb/linux-acpi-2.6: (39 commits)
ACPI: EC: Workaround for optimized controllers (version 3)
ACPI: EC: use printk_ratelimit(), add some DEBUG mode messages
Revert "ACPI: EC: Workaround for optimized controllers"
ACPI: fix two IRQ8 issues in IOAPIC mode
ACPI: Add missing spaces to printk format
cpuidle: fix HP nx6125 regression
cpuidle: add sched_clock_idle_[sleep|wakeup]_event() hooks
cpuidle: fix C3 for no bus-master control case
ACPI: thinkpad-acpi: fix oops when a module parameter has no value
Revert "Fix very high interrupt rate for IRQ8 (rtc) unless pnpacpi=off"
ACPI: EC: Don't init EC early if it has no _INI
Revert "acpi: make ACPI_PROCFS default to y"
Revert "ACPI: add documentation for deprecated /proc/acpi/battery in ACPI_PROCFS"
ACPI: Split out control for /proc/acpi entries from battery, ac, and sbs.
ACPI: Video: Increase buffer size for writes to brightness proc file.
ACPI: EC: Workaround for optimized controllers
ACPI: SBS: Fix retval warning
ACPI: Enable MSR (FixedHW) support for T-States
ACPI: Get throttling info from BIOS only after evaluating _PDC
ACPI: Use _TSS for throttling control, when present. Add error checks.
...
Linus Torvalds [Tue, 27 Nov 2007 03:05:19 +0000 (19:05 -0800)]
Merge branch 'upstream' of git://ftp.linux-mips.org/pub/scm/upstream-linus
* 'upstream' of git://ftp.linux-mips.org/pub/scm/upstream-linus:
[MIPS] vpe: Add missing "space"
[MIPS] Compliment va_start() with va_end().
[MIPS] IP22: Fix broken eeprom access by using __raw_readl/__raw_writel
[MIPS] IP22: Fix broken EISA interrupt setup by switching to generic i8259
[MIPS] 64-bit Sibyte kernels need DMA32.
[MIPS] Only build r4k clocksource for systems that work ok with it.
[MIPS] Handle R4000/R4400 mfc0 from count register.
[MIPS] Fix possible hang in LL/SC futex loops.
[MIPS] Fix context DSP context / TLS pointer switching bug for new threads.
[MIPS] IP32: More interrupt renumbering fixes.
[MIPS] time: MIPSsim's plat_time_init doesn't need to be irq safe.
[MIPS] time: Fix negated condition in cevt-r4k driver.
[MIPS] Fix pcspeaker build.
Andreas Herrmann [Mon, 19 Nov 2007 22:58:57 +0000 (23:58 +0100)]
x86: correctly set UTS_MACHINE for "make ARCH=x86"
For a kernel built with "make ARCH=x86" the following system
information is displayed when running the new kernel
$ uname -m
x86
On some i386 systems (e.g. K7) we even have the following information
$ uname -m
x66
This is weird. The usual information for "uname -m" should be "x86_64"
on 64-bit and "i386" or "i686" on 32-bit.
This patch fixes the issue by setting UTS_MACHINE to "i386" for 32-bit
kernel builds and to "x86_64" for 64-bit kernel builds. I.e., "x86"
won't be used for UTS_MACHINE anymore.
Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Ingo Molnar <mingo@redhat.com> Cc: "H. Peter Anvin" <hpa@zytor.com> Acked-by: Sam Ravnborg <sam@ravnborg.org> Signed-off-by: Andreas Herrmann <aherrman@arcor.de> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Russell King [Fri, 16 Nov 2007 22:13:24 +0000 (22:13 +0000)]
NFS: mount failure causes bad page state
While testing a kernel based upon ecd744eec3aa8bbc949ec04ed3fbf7ecb2958a0e
(with wrong boot arguments), I got the following bad page state entry while
NFS was trying to mount it's rootfs:
IP-Config: Complete:
device=eth0, addr=192.168.1.101, mask=255.255.255.0, gw=255.255.255.255,
host=192.168.1.101, domain=, nis-domain=(none),
bootserver=192.168.1.100, rootserver=192.168.1.100, rootpath=
Looking up port of RPC 100003/2 on 192.168.1.100
rpcbind: server 192.168.1.100 not responding, timed out
Root-NFS: Unable to get nfsd port number from server, using default
Looking up port of RPC 100005/1 on 192.168.1.100
rpcbind: server 192.168.1.100 not responding, timed out
Root-NFS: Unable to get mountd port number from server, using default
mount: server 192.168.1.100 not responding, timed out
Root-NFS: Server returned error -5 while mounting /nfs/rootfs/
VFS: Unable to mount root fs via NFS, trying floppy.
Bad page state in process 'swapper'
page:c02b1260 flags:0x00000400 mapping:00000000 mapcount:0 count:0
Trying to fix it up, but a reboot is needed
Backtrace:
[<c0023e34>] (dump_stack+0x0/0x14) from [<c0062570>] (bad_page+0x70/0xac)
[<c0062500>] (bad_page+0x0/0xac) from [<c0064914>] (free_hot_cold_page+0x80/0x178)
[<c0064894>] (free_hot_cold_page+0x0/0x178) from [<c0064a74>] (free_hot_page+0x14/0x18)
[<c0064a60>] (free_hot_page+0x0/0x18) from [<c0067078>] (put_page+0xf8/0x154)
[<c0066f80>] (put_page+0x0/0x154) from [<c007dbc8>] (kfree+0xc8/0xd0)
[<c007db00>] (kfree+0x0/0xd0) from [<c00cbb54>] (nfs_get_sb+0x230/0x710)
[<c00cb924>] (nfs_get_sb+0x0/0x710) from [<c0084334>] (vfs_kern_mount+0x58/0xac)[<c00842dc>] (vfs_kern_mount+0x0/0xac) from [<c00843c0>] (do_kern_mount+0x38/0xf4)
[<c0084388>] (do_kern_mount+0x0/0xf4) from [<c0099c7c>] (do_mount+0x1e8/0x614)
...
This seems to be caused by use of an uninitialised structure due to NULL
options being passed to nfs_validate_mount_data(). Ensure that the
parsed mount data is always initialised.
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
(Trond: added fix for the same bug in nfs4_validate_mount_data()). Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
James Lentini [Mon, 26 Nov 2007 17:42:44 +0000 (12:42 -0500)]
SUNRPC: remove NFS/RDMA client's binary sysctls
Support for binary sysctls is being deprecated in 2.6.24. Since there
are no applications using the NFS/RDMA client's binary sysctls, it
makes sense to remove them. The patch below does this while leaving
the /proc/sys interface unchanged.
Please consider this for 2.6.24.
Signed-off-by: James Lentini <jlentini@netapp.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Ingo Molnar [Mon, 26 Nov 2007 20:21:49 +0000 (21:21 +0100)]
sched: fix RLIMIT_CPU comment
Devan Lippman noticed that the RLIMIT_CPU comment in resource.h is
incorrect: the field is in seconds, not msecs. We used msecs in
earlier versions of the patch but that got changed.
Ingo Molnar [Mon, 26 Nov 2007 20:21:49 +0000 (21:21 +0100)]
sched: fix prev_stime calculation
Srivatsa Vaddagiri noticed occasionally incorrect CPU usage
values in top and tracked it down to stime going below 0 in
task_stime(). Negative values are possible there due to the
sampled nature of stime/utime.
Ingo Molnar [Mon, 26 Nov 2007 19:42:20 +0000 (20:42 +0100)]
x86: fix APIC related bootup crash on Athlon XP CPUs
warmbloodedcreature@gmail.com reported that an APIC-enabled
Asus a7v8x-x with an Athlon XP reboots early in the bootup:
http://bugzilla.kernel.org/show_bug.cgi?id=8723
after a long marathon of spontaneous-reboot debugging, it turns
out to be caused by sync_Arb_ids(). AMD CPUs never really needed
this sequence anyway, so just return early if we meet an AMD CPU.
Signed-off-by: Ingo Molnar <mingo@elte.hu> Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
John Stultz [Mon, 26 Nov 2007 19:42:19 +0000 (20:42 +0100)]
time: add ADJ_OFFSET_SS_READ
Michael Kerrisk reported that a long standing bug in the adjtimex()
system call causes glibc's adjtime(3) function to deliver the wrong
results if 'delta' is NULL.
add the ADJ_OFFSET_SS_READ API detail, which will be used by glibc
to fix this API compatibility bug.
Also see: http://bugzilla.kernel.org/show_bug.cgi?id=6761
[ mingo@elte.hu: added patch description and made it backwards compatible ]
NOTE: the new flag is defined 0xa001 so that it returns -EINVAL on
older kernels - this way glibc can use it safely. Suggested by Ulrich
Drepper.
Theodore Ts'o [Mon, 26 Nov 2007 19:42:19 +0000 (20:42 +0100)]
x86: export the symbol empty_zero_page on the 32-bit x86 architecture
The latest KVM driver wants to use the empty_zero_page symbol, and it's
not exported in 32-bit x86 (although it is exported by x86_64, s390, and
uml architectures).
Andrew Morton [Mon, 26 Nov 2007 19:42:19 +0000 (20:42 +0100)]
x86: fix kprobes_64.c inlining borkage
fix:
arch/x86/kernel/kprobes_64.c: In function 'set_current_kprobe':
arch/x86/kernel/kprobes_64.c:152: sorry, unimplemented: inlining failed in call to 'is_IF_modifier': recursive inlining
arch/x86/kernel/kprobes_64.c:166: sorry, unimplemented: called from here
Cc: Thomas Gleixner <tglx@linutronix.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Cc: mingo@elte.hu Cc: akpm@linux-foundation.org Cc: tglx@linutronix.de Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Signed-off-by: Ingo Molnar <mingo@elte.hu>
Michal Schmidt [Mon, 26 Nov 2007 19:42:19 +0000 (20:42 +0100)]
pci: use pci=bfsort for HP DL385 G2, DL585 G2
HP ProLiant systems DL385 G2 and DL585 G2 need pci=bfsort to enumerate PCI
devices in the expected order.
Matt sayeth:
biosdevname is a userspace app I wrote to help solve this so we don't need
to patch the kernel for future systems. It's not integrated into any
distributions properly yet, but is included in openSUSE 10.3 and Fedora 8
for people who want to download and install it there. It acts as a udev
helper.
For the time being, patching the kernel is necessary. I really hope
biosdevname eliminates that need in future distributions.
http://linux.dell.com/biosdevname/
Signed-off-by: Michal Schmidt <mschmidt@redhat.com> Acked-by: Andy Gospodarek <andy@greyhouse.net> Cc: mingo@elte.hu Cc: andy@greyhouse.net Cc: john.cagle@hp.com Cc: Matt Domsch <Matt_Domsch@dell.com> Cc: Greg KH <greg@kroah.com> Cc: Andi Kleen <ak@suse.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Signed-off-by: Ingo Molnar <mingo@elte.hu>
Andreas Herrmann [Mon, 26 Nov 2007 19:42:19 +0000 (20:42 +0100)]
x86: correctly set UTS_MACHINE for "make ARCH=x86"
x86: correctly set UTS_MACHINE for "make ARCH=x86"
For a kernel built with "make ARCH=x86" the following system
information is displayed when running the new kernel
$ uname -m
x86
On some i386 systems (e.g. K7) we even have the following information
$ uname -m
x66
This is weird. The usual information for "uname -m" should be "x86_64"
on 64-bit and "i386" or "i686" on 32-bit.
This patch fixes the issue by setting UTS_MACHINE to "i386" for 32-bit
kernel builds and to "x86_64" for 64-bit kernel builds. I.e., "x86"
won't be used for UTS_MACHINE anymore.
Acked-by: Sam Ravnborg <sam@ravnborg.org> Signed-off-by: Andreas Herrmann <aherrman@arcor.de> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: Roman Zippel <zippel@linux-m68k.org> Cc: Andrew Morton <akpm@linux-foundation.org> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Signed-off-by: Ingo Molnar <mingo@redhat.com>
Arjan van de Ven [Mon, 26 Nov 2007 19:42:19 +0000 (20:42 +0100)]
x86: printk kernel version in WARN_ON and other dump_stack users
today, all oopses contain a version number of the kernel, which is nice
because the people who actually do bother to read the oops get this
vital bit of information always without having to ask the reporter in
another round trip.
However, WARN_ON() and many other dump_stack() users right now lack this
information; the patch below adds this. This information is essential
for getting people to use their time effectively when looking at these
things; in addition, it's essential for tools that try to collect
statistics about defects.
Please consider, since its so simple and important for long term kernel
quality processes.
The code is identical between 32/64 bit; a lot of this code should be
unified over time, the patch keeps the identical-ness intact.
Signed-off-by: Arjan van de Ven <arjan@linux.intel.com> Signed-off-by: Ingo Molnar <mingo@elte.hu> Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
More than 3 years ago Niclas Gustafsson reported a 'stopped time'
problem:
> Watching the /proc/interrupts with 10s apart after the "stop".
>
> [root@s151 root]# more /proc/interrupts
> CPU0
> 0: 66413955 local-APIC-edge timer
[...]
> LOC: 67355837
> ERR: 0
> MIS: 0
> [root@s151 root]# more /proc/interrupts
> CPU0
> 0: 66413955 local-APIC-edge timer
[...]
> LOC: 67379568
> ERR: 0
> MIS: 0
This may be because buggy SMM firmware messes with the 8259A (configured
for a transparent mode -- yes that rare "local-APIC-edge" mode is tricky
;-) ) insanely.
Ralf Baechle [Sat, 3 Nov 2007 02:05:43 +0000 (02:05 +0000)]
[MIPS] 64-bit Sibyte kernels need DMA32.
Sibyte SOCs only have 32-bit PCI. Due to the sparse use of the address
space only the first 1GB of memory is mapped at physical addresses
below 1GB. If a system has more than 1GB of memory 32-bit DMA will
not be able to reach all of it.
For now this patch is good enough to keep Sibyte users happy but it seems
eventually something like swiotlb will be needed for Sibyte.
Ralf Baechle [Sat, 24 Nov 2007 22:33:28 +0000 (22:33 +0000)]
[MIPS] Only build r4k clocksource for systems that work ok with it.
In particular as-is it's not suited for multicore and mutiprocessors
systems where there is on guarantee that the counter are synchronized
or running from the same clock at all. This broke Sibyte and probably
others since the "[MIPS] Handle R4000/R4400 mfc0 from count register."
commit.
Ralf Baechle [Wed, 21 Nov 2007 16:39:44 +0000 (16:39 +0000)]
[MIPS] Handle R4000/R4400 mfc0 from count register.
The R4000 and R4400 have an errata where if the cp0 count register is read
in the exact moment when it matches the compare register no interrupt will
be generated.
This bug may be triggered if the cp0 count register is being used as
clocksource and the compare interrupt as clockevent. So a simple
workaround is to avoid using the compare for both facilities on the
affected CPUs.
This is different from the workaround suggested in the old errata documents;
at some opportunity probably the official version should be implemented
and tested. Another thing to find out is which processor versions
exactly are affected. I only have errata documents upto R4400 V3.0
available so for the moment the code treats all R4000 and R4400 as broken.
This is potencially a problem for some machines that have no other decent
clocksource available; this workaround will cause them to fall back to
another clocksource, worst case the "jiffies" source.
The branch at the end of the fixup code, it goes back to the SC
instruction, no matter if the fault was first taken by the LL or SC
instruction resulting in an endless loop which will only terminate if
the address become valid again due to another thread setting up an
accessible mapping and the CPU happens to execute the SC instruction
successfully which due to the preceeding ERET instruction of the fault
handler would only happen if UNPREDICTABLE instruction behaviour of the
SC instruction without a preceeding LL happens to favor that outcome.
But normally processes are nice, pass valid arguments and we were just
getting away with this.
Thanks to Kaz Kylheku <kaz@zeugmasystems.com> for providing the original
report and a test case.
Ralf Baechle [Tue, 30 Oct 2007 17:25:26 +0000 (17:25 +0000)]
[MIPS] Fix context DSP context / TLS pointer switching bug for new threads.
A new born thread starts execution not in schedule but rather in
ret_from_fork which results in it bypassing the part of the code to
load a new context written in C which are the DSP context and the
userlocal register which Linux uses for the TLS pointer. Frequently
we were just getting away with this bug for a number of reasons:
o Real world application scenarios are very unlikely to use clone or fork
in blocks of DSP code.
o Linux by default runs the child process right after the fork, so the
child by luck will find all the right context in the DSP and userlocal
registers.
o So far the rdhwr instruction was emulated on all hardware so userlocal
wasn't getting referenced at all and the emulation wasn't suffering
from the issue since it gets it's value straight from the thread's
thread_info.
Fixed by moving the code to load the context from switch_to() to
finish_arch_switch which will be called by newborn and old threads.
Tejun Heo [Mon, 26 Nov 2007 11:58:02 +0000 (20:58 +0900)]
libata: bump transfer chunk size if it's odd
None of the drives I have follows what the standard says about
transfer chunk size. Of the four SATA and six PATA ATAPI devices
tested, four ignore transfer chunk size completely and the ones which
honor it don't behave according to the spec when it's odd.
According to the spec, transfer chunk size can be odd if the amount of
data to transfer equals or is smaller than the chunk size and the
device can indicate the same odd number and transfer the whole thing
at one go with a pad byte appended. However, in reality, none of the
drives I have does that. They all indicate and transfer even number
of bytes one byte shorter than the chunk size first; then indicate and
transfer two bytes, which is clearly out of spec.
In addition to unnecessary second PIO data phase, this also creates a
weird problem when combined with SATA controllers which perform PIO
via DMA. Some of these controllers use actualy number of bytes
received to update DMA pointer so chunks which are sized 4n + 2 makes
DMA pointer off by two bytes. This causes data corruption and buffer
overruns.
This patch rounds nbytes up to the nearest even number such that ATAPI
devices don't split data transfer for the last odd byte. This
shouldn't confuse controllers which depend on transfer chunk size as
devices will report the rounded-up number, actually transfer that much
and padding buffer is there to receive them.
Signed-off-by: Tejun Heo <htejun@gmail.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
Adrian Bunk [Mon, 26 Nov 2007 15:17:53 +0000 (23:17 +0800)]
[IPV4]: Remove bogus ifdef mess in arp_process
The #ifdef's in arp_process() were not only a mess, they were also wrong
in the CONFIG_NET_ETHERNET=n and (CONFIG_NETDEV_1000=y or
CONFIG_NETDEV_10000=y) cases.
Since they are not required this patch removes them.
Also removed are some #ifdef's around #include's that caused compile
errors after this change.
Signed-off-by: Adrian Bunk <bunk@kernel.org> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Herbert Xu [Mon, 26 Nov 2007 15:11:19 +0000 (23:11 +0800)]
[SKBUFF]: Free old skb properly in skb_morph
The skb_morph function only freed the data part of the dst skb, but leaked
the auxiliary data such as the netfilter fields. This patch fixes this by
moving the relevant parts from __kfree_skb to skb_release_all and calling
it in skb_morph.
It also makes kfree_skbmem static since it's no longer called anywhere else
and it now no longer does skb_release_data.
Thanks to Yasuyuki KOZAKAI for finding this problem and posting a patch for
it.
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
The error is obvious - if the NUMA is on and the size
is less than the PAGE_SIZE we leak the pointer (kfree is
inside the #else branch).
Compiler doesn't warn us because after the kfree(x) there's
a "x = NULL" assignment, so here's another (minor?) bug: we
don't set x to NULL under certain circumstances.
Boring explanation, I know... Patch explains it better.
Signed-off-by: Pavel Emelyanov <xemul@openvz.org> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
[IPSEC]: Lock state when copying non-atomic fields to user-space
caused a regression.
Ingo Molnar reports that it causes a potential dead-lock found by the
lock validator as it tries to take x->lock within xfrm_state_lock while
numerous other sites take the locks in opposite order.
For 2.6.24, the best fix is to simply remove the added locks as that puts
us back in the same state as we've been in for years. For later kernels
a proper fix would be to reverse the locking order for every xfrm state
user such that if x->lock is taken together with xfrm_state_lock then
it is to be taken within it.
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Tejun Heo [Thu, 22 Nov 2007 09:46:57 +0000 (18:46 +0900)]
sata_sil24: fix sg table sizing
sil24 unnecessarily used LIBATA_MAX_PRD and ATAPI sg table was short
by one entry which might cause very obscure problems. This patch
updates sg table sizing such that
* One full page is used for PRB + sg table. On 4k page,
this results in 253 sg's.
* Make ATAPI sg block properly sized.
* Make build fail if command block size doesn't equal PAGE_SIZE.
Signed-off-by: Tejun Heo <htejun@gmail.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
Tejun Heo [Mon, 19 Nov 2007 09:06:11 +0000 (18:06 +0900)]
pata_jmicron: fix disabled port handling in jmicron_pre_reset()
There are two bugs in disabled port handling.
* test in PORT_PATA0 is reversed
* ->prereset should return -ENOENT for disabled ports not 0
The first bug makes the PATA channel considered disabled but the
second bug saves the day by returning 0. The net result is that cable
is always left at ATA_CBL_UNKNOWN. This results in false 80c
configuration and thus transfer errors.
This patch fixes both bugs.
Signed-off-by: Tejun Heo <htejun@gmail.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
Herbert Xu [Sat, 17 Nov 2007 14:09:38 +0000 (22:09 +0800)]
[CRYPTO] api: Fix potential race in crypto_remove_spawn
As it is crypto_remove_spawn may try to unregister an instance which is
yet to be registered. This patch fixes this by checking whether the
instance has been registered before attempting to remove it.
It also removes a bogus cra_destroy check in crypto_register_instance as
1) it's outside the mutex;
2) we have a check in __crypto_register_alg already.
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Herbert Xu [Thu, 11 Oct 2007 08:45:17 +0000 (16:45 +0800)]
[CRYPTO] authenc: Move initialisations up to shut up gcc
It seems that newer versions of gcc have regressed in their abilities to
analyse initialisations. This patch moves the initialisations up to avoid
the warnings.
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Ilpo Järvinen [Fri, 23 Nov 2007 11:10:56 +0000 (19:10 +0800)]
[TCP] MTUprobe: Cleanup send queue check (no need to loop)
The original code has striking complexity to perform a query
which can be reduced to a very simple compare.
FIN seqno may be included to write_seq but it should not make
any significant difference here compared to skb->len which was
used previously. One won't end up there with SYN still queued.
Use of write_seq check guarantees that there's a valid skb in
send_head so I removed the extra check.
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi> Acked-by: John Heffner <jheffner@psc.edu> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Ilpo Järvinen [Fri, 23 Nov 2007 11:08:16 +0000 (19:08 +0800)]
[TCP]: MTUprobe: receiver window & data available checks fixed
It seems that the checked range for receiver window check should
begin from the first rather than from the last skb that is going
to be included to the probe. And that can be achieved without
reference to skbs at all, snd_nxt and write_seq provides the
correct seqno already. Plus, it SHOULD account packets that are
necessary to trigger fast retransmit [RFC4821].
Location of snd_wnd < probe_size/size_needed check is bogus
because it will cause the other if() match as well (due to
snd_nxt >= snd_una invariant).
Removed dead obvious comment.
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Magnus Damm [Fri, 23 Nov 2007 05:39:13 +0000 (14:39 +0900)]
sh: fix R2D-1 CF support
This patch fixes CF support for R2D-1 boards. Both R2D-1 and
R2D-PLUS are equipped with CF IRQs, but the R2D-1 FPGA version
seem to deliver IRQ spikes with certain CF cards during libata
probing. This patch enables polling for R2D-1 as a workaround
for this broken FGPA logic.
Is being held until the list moderator can review it for approval.
The reason it is being held:
Post by non-member to a members-only list
Signed-off-by: Gabriel Craciunescu <nix.or.die@googlemail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Adrian Bunk [Thu, 22 Nov 2007 11:38:25 +0000 (19:38 +0800)]
[SUNRPC]: Make xprtsock.c:xs_setup_{udp,tcp}() static
xs_setup_{udp,tcp}() can now become static.
Signed-off-by: Adrian Bunk <bunk@kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Pavel Emelyanov [Thu, 22 Nov 2007 11:15:56 +0000 (19:15 +0800)]
[IRDA]: Compilation for CONFIG_INET=n case
Found this occasionally.
The CONFIG_INET=n is hardly ever set, but if it is the
irlan_eth_send_gratuitous_arp() compilation should produce a
warning about unused variable in_dev.
Too pedantic? :)
Signed-off-by: Pavel Emelyanov <xemul@openvz.org> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
David Woodhouse [Wed, 21 Nov 2007 17:45:12 +0000 (18:45 +0100)]
mmc: Avoid re-using minor numbers before the original device is closed.
Move the code which marks the minor number as free to mmc_blk_put() so
that it happens on the final close() (or removal), instead of doing it
at removal even when the device is still logically open.
Signed-off-by: David Woodhouse <dwmw2@infradead.org> Signed-off-by: Pierre Ossman <drzeus@drzeus.cx>
Alex Dubov [Wed, 14 Nov 2007 12:55:36 +0000 (23:55 +1100)]
tifm_sd: handle non-power-of-2 block sizes
It is possible to handle arbitrary block sizes with tifm card reader by
conditionally switching to PIO in case such block has to be delivered. At
the beginning of each request, DMA is either disabled (non-power-of-2 block
size) or set to load time user preference.
Signed-off-by: Alex Dubov <oakad@yahoo.com> Signed-off-by: Pierre Ossman <drzeus@drzeus.cx>
projects / linux-2.6 / log