Joachim Fenkes [Wed, 29 Aug 2007 16:15:17 +0000 (18:15 +0200)]
[POWERPC] ibmebus: Prevent bus_id collisions
Previously, ibmebus derived a device's bus_id from its location code.
The location code is not guaranteed to be unique, so we might get bus_id
collisions if two devices share the same location code. The OFDT
full_name, however, is unique, so we use that instead (truncating it
on the left if it is too long).
Signed-off-by: Joachim Fenkes <fenkes@de.ibm.com> Signed-off-by: Paul Mackerras <paulus@samba.org>
[POWERPC] cell/PS3: Ignore storage devices that are still being probed
On PS3, A storage device may show up in the repository before the hypervisor
has finished probing:
- If its type is not yet known, it shows up as PS3_DEV_TYPE_STOR_DUMMY,
- If its regions are being probed, it shows up as having zero regions.
If any of these happen, consider the device not yet present. The storage
probe thread will retry later.
This fixes the timing-dependent problem where a kernel booted from FLASH ROM
sometimes cannot find the hard disk.
Jeremy Kerr [Fri, 7 Sep 2007 08:28:27 +0000 (18:28 +1000)]
[POWERPC] cell/PS3: Always set master run control bit in mfc_sr1_set
At present, running any SPE program on the ps3 will trigger a BUG_ON
when spufs_run_spu tries to clear the master run control bit, as lv1
does not make the master run control available to Linux.
This change makes SPE apps work again by disabling changes to the
master run control on PS3. Although we don't have the facility to
disable a SPE with supervisor-level privileges, it's better than
hitting the BUG_ON unconditionally.
Signed-off-by: Jeremy Kerr <jk@ozlabs.org> Acked-by: Masato Noguchi <Masato.Noguchi@jp.sony.com> Signed-off-by: Paul Mackerras <paulus@samba.org>
[POWERPC] cell/PS3: Fix a bug that causes the PS3 to hang on the SPU Class 0 interrupt.
The Cell BE Architecture spec states that the SPU MFC Class 0 interrupt
is edge-triggered. The current spu interrupt handler assumes this
behavior and does not clear the interrupt status.
The PS3 hypervisor visualizes all SPU interrupts as level, and on return
from the interrupt handler the hypervisor will deliver a new virtual
interrupt for any unmasked interrupts which for which the status has not
been cleared. This fix clears the interrupt status in the interrupt
handler.
Signed-off-by: Masato Noguchi <Masato.Noguchi@jp.sony.com> Signed-off-by: Geoff Levand <geoffrey.levand@am.sony.com> Signed-off-by: Jeremy Kerr <jk@ozlabs.org> Acked-by: Arnd Bergmann <arnd.bergmann@de.ibm.com> Signed-off-by: Paul Mackerras <paulus@samba.org>
Herbert Xu [Sun, 9 Sep 2007 07:45:21 +0000 (08:45 +0100)]
[CRYPTO] blkcipher: Fix handling of kmalloc page straddling
The function blkcipher_get_spot tries to return a buffer of
the specified length that does not straddle a page. It has
an off-by-one bug so it may advance a page unnecessarily.
What's worse, one of its callers doesn't provide a buffer
that's sufficiently long for this operation.
This patch fixes both problems. Thanks to Bob Gilligan for
diagnosing this problem and providing a fix.
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Stefan Richter [Thu, 6 Sep 2007 07:50:30 +0000 (09:50 +0200)]
firewire: fw-ohci: ignore failure of pci_set_power_state (fix suspend regression)
Fixes (papers over) "Sleep problems with kernels >= 2.6.21 on powerpc",
http://lkml.org/lkml/2007/8/25/155. The issue is that the FireWire
controller's pci_dev.current_state of iBook G3 and presumably older
PowerBooks is still in PCI_UNKNOWN instead of PCI_D0 when the firewire
driver's .suspend method is called.
Like it was suggested earlier in http://lkml.org/lkml/2006/10/24/13, we
do not fail .suspend anymore if pci_set_power_state failed.
Signed-off-by: Stefan Richter <stefanr@s5r6.in-berlin.de>
Hirokazu Takata [Mon, 20 Aug 2007 11:53:50 +0000 (20:53 +0900)]
m32r: Rename STI/CLI macros
The names of STI and CLI macros were derived from i386 arch historically,
but their name are incomprehensible.
So, for easy to understand, rename these macros to ENABLE_INTERRUPTS
and DISABLE_INTERRUPTS, respectively.
Peter Zijlstra [Wed, 5 Sep 2007 12:32:49 +0000 (14:32 +0200)]
sched: improve prev_sum_exec_runtime setting
Second preparatory patch for fix-ideal runtime:
Mark prev_sum_exec_runtime at the beginning of our run, the same spot
that adds our wait period to wait_runtime. This seems a more natural
location to do this, and it also reduces the code a bit:
text data bss dec hex filename
13397 228 1204 14829 39ed sched.o.before
13391 228 1204 14823 39e7 sched.o.after
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Signed-off-by: Ingo Molnar <mingo@elte.hu>
fix niced_granularity(). This resulted in under-scheduling for
CPU-bound negative nice level tasks (and this in turn caused
higher than necessary latencies in nice-0 tasks).
sched: fix MC/HT scheduler optimization, without breaking the FUZZ logic.
First fix the check
if (*imbalance + SCHED_LOAD_SCALE_FUZZ < busiest_load_per_task)
with this
if (*imbalance < busiest_load_per_task)
As the current check is always false for nice 0 tasks (as
SCHED_LOAD_SCALE_FUZZ is same as busiest_load_per_task for nice 0
tasks).
With the above change, imbalance was getting reset to 0 in the corner
case condition, making the FUZZ logic fail. Fix it by not corrupting the
imbalance and change the imbalance, only when it finds that the HT/MC
optimization is needed.
Eric Sandeen [Thu, 16 Aug 2007 06:49:11 +0000 (16:49 +1000)]
[XFS] fix nasty quota hashtable allocation bug
This git mod: 77e4635ae191774526ed695482a151ac986f3806
converted to a "greedy" allocation interface, but for the quota hashtables
it switched from allocating XFS_QM_HASHSIZE (nr of elements)
xfs_dqhash_t's to allocating only XFS_QM_HASHSIZE *bytes* - quite a lot
smaller! Then when we converted hsize "back" to nr of elements (the
division line) hsize went to 0. This was leading to oopses when running
any quota tests on the Fedora 8 test kernel, but the problem has been
there for almost a year.
- in xfs_probe_cluster rename the inner len to pg_len. There's no harm
here because the outer len isn't used after the inner len comes into
existence but it keeps the code clean.
- in xfs_da_do_buf remove the inner i because they don't overlap
and they are both the same type.
- remove the != 0 inside the unlikely in ASSERT_ALWAYS because sparse now
complains about comparisons between pointers and 0
- add a standalone ASSERT implementation because defining it to
ASSERT_ALWAYS means the string is expanded before the token passing
stringification. This way we get the actual content of the
assertion in the assfail message and don't overflow sparse's
stringification buffer leading to sparse error messages.
Apparently XEN does not keep the contents of the 48-bit gdt_48 data
structure that is passed to lgdt in the XEN machine state. Instead it
appears to save the _address_ of the 48-bit descriptor
somewhere. Unfortunately this data happens to reside on the stack and
is probably no longer availiable at the time of the actual protected
mode jump.
This is Xen bug but given that there is a one-line patch to work
around this problem, the linux kernel should probably do this. My fix
is to make the gdt_48 description in setup_gdt static (in setup_idt
this is already the case). This allows the kernel to boot under
Xen HVM again.
Signed-off-by: Christian Ehrhardt <lk@c--e.de> Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Hirokazu Takata [Mon, 20 Aug 2007 00:12:46 +0000 (09:12 +0900)]
m32r: Separate syscall table from entry.S
- Separate sys_call_table from arch/m32r/kernel/entry.S and
move it to arch/m32r/kernel/system_call.S.
- Change sys_call_table section from .data to .rodata.
Hirokazu Takata [Fri, 17 Aug 2007 09:11:37 +0000 (18:11 +0900)]
m32r: Simplify ei_handler code
Simplify and clean up messy ei_handler code in arch/m32r/kernel/entry.S.
- Remove ifdef's for CONFIG_CHIP_* configulations.
- Rearrange the M32700 workaround code.
- Remove the messy platform-dependent interrupt check routines and
consolidate them to common INT0/INT1/INT2 check routines for all
platforms with cascaded interrupt controllers.
Ensure pages are uptodate after returning from read_cache_page, which allows
us to cut out most of the filesystem-internal PageUptodate calls.
I didn't have a great look down the call chains, but this appears to fixes 7
possible use-before uptodate in hfs, 2 in hfsplus, 1 in jfs, a few in
ecryptfs, 1 in jffs2, and a possible cleared data overwritten with readpage in
block2mtd. All depending on whether the filler is async and/or can return
with a !uptodate page.
It introduced a wait to read_cache_page, as well as a
read_cache_page_async function equivalent to the old read_cache_page
without any callers.
Switching jffs2_gc_fetch_page to read_cache_page_async for the old
behavior makes the deadlocks go away, but maybe reintroduces the
use-before-uptodate problem? I don't understand the mm/fs interaction
well enough to say.
[It's fine. dwmw2.]
Signed-off-by: Jason Lunz <lunz@falooley.org> Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Trond Myklebust [Tue, 28 Aug 2007 14:29:36 +0000 (10:29 -0400)]
NFS: Fix a write request leak in nfs_invalidate_page()
Ryusuke Konishi says:
The recent truncate_complete_page() clears the dirty flag from a page
before calling a_ops->invalidatepage(),
^^^^^^
static void
truncate_complete_page(struct address_space *mapping, struct page *page)
{
...
cancel_dirty_page(page, PAGE_CACHE_SIZE); <--- Inserted here at
kernel 2.6.20
if (PagePrivate(page))
do_invalidatepage(page, 0); ---> will call
a_ops->invalidatepage()
...
}
and this is disturbing nfs_wb_page_priority() from calling
nfs_writepage_locked() that is expected to handle the pending
request (=nfs_page) associated with the page.
int nfs_wb_page_priority(struct inode *inode, struct page *page, int how)
{
...
if (clear_page_dirty_for_io(page)) {
ret = nfs_writepage_locked(page, &wbc);
if (ret < 0)
goto out;
}
...
}
Since truncate_complete_page() will get rid of the page after
a_ops->invalidatepage() returns, the request (=nfs_page) associated
with the page becomes a garbage in nfs_inode->nfs_page_tree.
------------------------
Fix this by ensuring that nfs_wb_page_priority() recognises that it may
also need to clear out non-dirty pages that have an nfs_page associated
with them.
Chuck Lever [Wed, 29 Aug 2007 21:59:03 +0000 (17:59 -0400)]
NFS: change NFS mount error return when hostname/pathname too long
According to the mount(2) man page, the proper error return code for the
mount(2) system call when the special device name or the mounted-on
directory name is too long is ENAMETOOLONG.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Chuck Lever [Wed, 29 Aug 2007 21:58:59 +0000 (17:58 -0400)]
NFS: Return a real error code from mount(2)
Don't filter the return code from the in-kernel rpcbind or NFS mount
clients. Return the real error code so that callers of the new NFS
text-based mount API can apply a useful retry strategy.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Chuck Lever [Wed, 29 Aug 2007 21:58:57 +0000 (17:58 -0400)]
NFS: mount option parser chokes on proto=
The new text-based NFS mount option parsing logic doesn't recognize any
valid transport protocols due to a silly mistake in the protocol token
matching logic. This prevents basic mount requests such as:
mount.nfs server:/export /mnt -o proto=tcp
from working with the new text-based NFS mount API.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Ondrej Zary [Fri, 31 Aug 2007 22:55:21 +0000 (00:55 +0200)]
Fix sata_via write errors on PATA drive connected to VT6421
I think that I've found and fixed the problem. There is a copy/paste bug in
vt6421_set_dma_mode() function which causes wrong values to be written to
PATA_UDMA_TIMING register.
This patch fixes a copy/paste bug that breaks DMA modes on VT6421 PATA port.
Signed-off-by: Ondrej Zary <linux@rainbow-software.org> Acked-by: Alan Cox <alan@redhat.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
Peter Chubb [Tue, 21 Aug 2007 03:57:01 +0000 (13:57 +1000)]
[IA64] Cleanup HPSIM code (was: Re: Enable early console for Ski simulator)
After my last patch we have a new header file for HP simulator use.
Here's code to use it for stuff that used to have `extern' statements
inline in the code. Functionality should not change with this patch.
Signed-off-by: Peter Chubb <peterc@gelato.unsw.edu.au> Signed-off-by: Tony Luck <tony.luck@intel.com>
Peter Chubb [Tue, 21 Aug 2007 03:51:45 +0000 (13:51 +1000)]
[IA64] Enable early console for Ski simulator
This patch cleans up the `enable early console for SKI' patch
(471e7a44848f467c9b83adc3463d019d2fa8817f), and
1. potentially allows the gensparse_defconfig to work again.
(there are other problems running a generic kernel on Ski)
2. fixes the `console registered twice' problem.
3. Cleans up the code by moving the `extern hpsim_cons' declaration to
a new asm/hpsim.h file.
Thanks to Jes for comments.
Signed-off-by: Peter Chubb <peterc@gelato.unsw.edu.au> Signed-off-by: Tony Luck <tony.luck@intel.com>
Prarit Bhargava [Wed, 22 Aug 2007 18:34:38 +0000 (14:34 -0400)]
[IA64] Stop bogus NMI & softlockup warnings in ia64 show_mem
When dumping memory via sysrq-m it is possible to take a bogus NMI watchdog
or softlockup watchdog because the dump can take a long time on big memory
systems.
Occasionally tickle the watchdog when doing the dump.
Signed-off-by: Prarit Bhargava <prarit@redhat.com> Signed-off-by: Tony Luck <tony.luck@intel.com>
John Keller [Thu, 23 Aug 2007 00:32:06 +0000 (19:32 -0500)]
[IA64] SN: Add support for CPU disable
Add additional support for CPU disable on SN platforms.
Correctly setup the smp_affinity mask for I/O error IRQs.
Restrict the use of the feature to Altix 4000 and 450 systems
running with a CPU disable capable PROM, and do not allow disabling
of CPU 0.
Signed-off-by: John Keller <jpk@sgi.com> Signed-off-by: Tony Luck <tony.luck@intel.com>
Roman Zippel [Thu, 30 Aug 2007 03:06:17 +0000 (05:06 +0200)]
kconfig: oldconfig shall not set symbols if it does not need to
Avoid setting the value if the symbol doesn't need to be changed or can't
be changed. Later choices may change the dependencies and thus the
possible input range.
make oldconfig from a 2.6.22 .config with CONFIG_HOTPLUG_CPU not set
was in some configurations setting CONFIG_HOTPLUG_CPU=y without asking,
even when there was no actual requirement for CONFIG_HOTPLUG_CPU.
This was triggered by SUSPEND_SMP that does a select HOTPLUG_CPU.
Signed-off-by: Roman Zippel <zippel@linux-m68k.org> Tested-by: Hugh Dickins <hugh@veritas.com> Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband:
IB/ehca: SRQ fixes to enable IPoIB CM
IB/ehca: Fix Small QP regressions
Trond Myklebust [Fri, 31 Aug 2007 14:45:17 +0000 (10:45 -0400)]
NFS: Fix the mount regression
This avoids the recent NFS mount regression (returning EBUSY when
mounting the same filesystem twice with different parameters).
The best I can do given the constraints appears to be to have the kernel
first look for a superblock that matches both the fsid and the
user-specified mount options, and then spawn off a new superblock if
that search fails.
Note that this is not the same as specifying nosharecache everywhere
since nosharecache will never attempt to match an existing superblock.
x86: be even more careful about checking the stack frame on dumping
lguest didn't initialize the kernel stack the way a real i386 kernel
does, and ended up triggering a corner-case in the stack frame checking
that doesn't happen on naive i386, and that the stack dumping didn't
handle quite right.
This makes the frame handling more correct, and tries to clarify the
code at the same time so that it's a bit more obvious what is going on.
Thanks to Rusty Russell for debugging the lguest failure-
Cc: Rusty Russell <rusty@rustcorp.com.au> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Do not use the ia64 clocksource on non-ia64 architectures
The HPET clocksource in drivers/char/hpet.c was written as generic code
for ia64, but it is not yet ready to replace the native HPET clocksource
implementations that the i386/x86-64 architectures use.
On x86[-64], trying to register this clocksource results in potentially
multiple hpet-based clocksources being registered, and if the ia64 one
is chosen on x86_64 some users have experienced hangs.
Eventually all three architectures may end up using the same code, but
that is not the case right now.
Cc: John Stultz <johnstul@us.ibm.com> Cc: Tony Luck <tony.luck@intel.com> Cc: Paolo Ornati <ornati@fastwebnet.it> Cc: Bob Picco <bob.picco@hp.com> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Merge branch 'upstream-linus' of master.kernel.org:/pub/scm/linux/kernel/git/jgarzik/netdev-2.6
* 'upstream-linus' of master.kernel.org:/pub/scm/linux/kernel/git/jgarzik/netdev-2.6:
- cxgb3 engine microcode load
cxgb3 - Fix dev->priv usage
qeth: Drop ARP packages on HiperSockets interface with NOARP attribute.
qeth: provide specific message for OSA-adapters exclusively used
qeth: crash during reboot after failing online setting
qeth: Announce tx checksumming for qeth devices in TSO/EDDP mode
qeth: dont return the return values of void functions.
qeth: enforce a rate limit for inbound scatter gather messages
qeth: ungrouping a device must not be interruptible
netxen: fix crashes during module unload
netxen: Avoid firmware load in PCI probe
PS3: fix the bug that 'ifconfig down' would hang
IOC3: Program UART predividers.
Joachim Fenkes [Fri, 31 Aug 2007 14:03:37 +0000 (16:03 +0200)]
IB/ehca: SRQ fixes to enable IPoIB CM
Fix ehca SRQ support so that IPoIB connected mode works:
- Report max_srq > 0 if SRQ is supported
- Report "last wqe reached" asynchronous event when base QP dies;
this is required by the IB spec and IPoIB CM relies on receiving it
when cleaning up.
Signed-off-by: Joachim Fenkes <fenkes@de.ibm.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
H. Peter Anvin [Fri, 31 Aug 2007 18:30:31 +0000 (11:30 -0700)]
[x86 setup] Don't rely on the VESA BIOS being register-clean
The VESA BIOS is specified to be register-clean. However, we have now
found at least one system which violates that. Thus, be as paranoid
about VESA calls as about everything else.
Huge thanks to Will Simoneau for reporting, diagnosing, and testing
this out on Dell Inspiron 5150.
Cc: Will Simoneau <simoneau@ele.uri.edu> Signed-off-by: H. Peter Anvin <hpa@zytor.com>
* git://git.kernel.org/pub/scm/linux/kernel/git/mingo/linux-2.6-sched:
sched: clean up task_new_fair()
sched: small schedstat fix
sched: fix wait_start_fair condition in update_stats_wait_end()
sched: call update_curr() in task_tick_fair()
sched: make the scheduler converge to the ideal latency
sched: fix sleeper bonus limit
Linus Torvalds [Fri, 31 Aug 2007 17:45:06 +0000 (10:45 -0700)]
Merge branch 'upstream-linus' of master.kernel.org:/pub/scm/linux/kernel/git/jgarzik/libata-dev
* 'upstream-linus' of master.kernel.org:/pub/scm/linux/kernel/git/jgarzik/libata-dev:
[libata] Bump driver versions
ata_piix: implement IOCFG bit18 quirk
libata: implement BROKEN_HPA horkage and apply it to affected drives
sata_promise: FastTrack TX4200 is a second-generation chip
pata_marvell: Add more identifiers
ata_piix: add Satellite U200 to broken suspend list
ata: add ATA_MWDMA* and ATA_SWDMA* defines
ata_piix: IDE mode SATA patch for Intel Tolapai
libata-core: Allow translation setting to fail
Divy Le Ray [Thu, 30 Aug 2007 02:15:52 +0000 (19:15 -0700)]
- cxgb3 engine microcode load
Load the engine microcode when an interface
is brought up, instead of of doing it when the module
is loaded.
Loosen up tight binding between the driver and the
engine microcode version.
There is no need for microcode update with T3A boards.
Fix the file naming.
Do a better job at logging the loading activity.
Signed-off-by: Divy Le Ray <divy@chelsio.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
Divy Le Ray [Thu, 30 Aug 2007 02:15:47 +0000 (19:15 -0700)]
cxgb3 - Fix dev->priv usage
cxgb3 used netdev_priv() and dev->priv for different purposes.
In 2.6.23, netdev_priv() == dev->priv, cxgb3 needs a fix.
This patch is a partial backport of Dave Miller's changes in the
net-2.6.24 git branch.
Without this fix, cxgb3 crashes on 2.6.23.
Signed-off-by: Divy Le Ray <divy@chelsio.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
Klaus D. Wacker [Wed, 29 Aug 2007 09:26:58 +0000 (11:26 +0200)]
qeth: Drop ARP packages on HiperSockets interface with NOARP attribute.
A network interface can get ARP packets even when the interface has
NOARP specified. In a HiperSockets environment this disturbs receiving
systems when packets are sent on the multicast queue. (E.g. TCP/IP on
z/VM issues messages reporting invalid data on the HiperSockets
interface.)
Qeth will no longer send ARP packets on HiperSockets interface when
interface has the NOARP attribute.
Signed-off-by: Klaus D. Wacker <kdwacker@de.ibm.com> Signed-off-by: Ursula Braun <braunu@de.ibm.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
Ursula Braun [Wed, 29 Aug 2007 09:26:57 +0000 (11:26 +0200)]
qeth: provide specific message for OSA-adapters exclusively used
Exclusive usage of OSA-cards has been introduced. Even though Linux
does not make use of it, qeth should be prepared to receive a bad RC
for some initialization steps. A meaningful message is now given,
if an OSA-device is set online, even though the OSA-adapter is already
exclusively used by another host.
Signed-off-by: Ursula Braun <braunu@de.ibm.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
Ursula Braun [Wed, 29 Aug 2007 09:26:56 +0000 (11:26 +0200)]
qeth: crash during reboot after failing online setting
Online setting of a qeth device may fail for instance because of:
- out-of-memory condition when allocating qdio queues
- IDX ACTIVATE problem
- ...
Such a device is still returned in a driver_for_each_device loop
processed in qeth_reboot_event(), which calls
qeth_clear_qdio_buffers(). Make sure qeth_clear_output_buffer() is
called only, if the qdio queues have been successfully allocated
during initialization of a qeth device.
Signed-off-by: Ursula Braun <braunu@de.ibm.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
Ursula Braun [Wed, 29 Aug 2007 09:26:52 +0000 (11:26 +0200)]
qeth: ungrouping a device must not be interruptible
Problem:
A recovery thread must not be active when device is removed.
In qeth_remove_device() an interruptible wait operation is used
to wait until a qeth recovery thread is finished. If a user really
interrupts the ungroup operation of a qeth device while a recovery
is running, cio and qeth are out of sync (device already removed
from cio, but kept in qeth). A following module unload of qeth
results in a kernel OOPS here.
Solution:
Do not allow interruption of ungroup operation to guarantee
finishing of a potentially running qeth recovery thread.
Signed-off-by: Ursula Braun <braunu@de.ibm.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
This patch fixes two problems during driver unload. The pci_disable_device()
call is before firmware reload, causing reads and writes across PCI bus after
disabling device. Second problem is the register window was wrong during
firmware reload
Loading firmware during PCI probe can lead to incorrect initialization,
rendering the card unusable until next reboot. This was introduced a while
ago as a workaround for firmware bug, a better workaround was submitted for
this a while ago. So removing original hack that loads firmware during probe.
Masakazu Mokuno [Tue, 28 Aug 2007 04:18:15 +0000 (13:18 +0900)]
PS3: fix the bug that 'ifconfig down' would hang
Fix the bug that 'ifconfig eth0 down' would hang up, reported by Stefan
Assmann <sassmann@suse.de>.
As we removed netif_poll_enable() from dev->open(), we should not use
netif_poll_disable() in dev->stop().
Ralf Baechle [Sun, 26 Aug 2007 17:51:22 +0000 (18:51 +0100)]
IOC3: Program UART predividers.
The IOC3 driver's UART detection bits used to rely on the the firmware
setting the UART pre-divider in a way that's apropriate for the 8250
driver which doesn't currently program this register. This happens
to work for the console but not rarely for additional ports.
While at it, also program the UART to RS-232 PIO mode; it the UART might
have been in mac-serial and/or DMA mode though that hasn't actually been
observed in practice.
Signed-off-by: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
Page migration: Do not accept invalid nodes in the target nodeset
Page migration currently does not check if the target of the move contains
nodes that that are invalid (if root attempts to migrate pages)
and may try to allocate from invalid nodes if these are specified
leading to oopses.
Return -EINVAL if an offline node is specified.
Signed-off-by: Christoph Lameter <clameter@sgi.com> Cc: Shaohua Li <shaohua.li@intel.com> Cc: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
David Gibson [Fri, 31 Aug 2007 06:56:40 +0000 (23:56 -0700)]
hugepage: fix broken check for offset alignment in hugepage mappings
For hugepage mappings, the file offset, like the address and size, needs to
be aligned to the size of a hugepage.
In commit 68589bc353037f233fe510ad9ff432338c95db66, the check for this was
moved into prepare_hugepage_range() along with the address and size checks.
But since BenH's rework of the get_unmapped_area() paths leading up to
commit 4b1d89290b62bb2db476c94c82cf7442aab440c8, prepare_hugepage_range()
is only called for MAP_FIXED mappings, not for other mappings. This means
we're no longer ever checking for an aligned offset - I've confirmed that
mmap() will (apparently) succeed with a misaligned offset on both powerpc
and i386 at least.
This patch restores the check, removing it from prepare_hugepage_range()
and putting it back into hugetlbfs_file_mmap(). I'm putting it there,
rather than in the get_unmapped_area() path so it only needs to go in one
place, than separately in the half-dozen or so arch-specific
implementations of hugetlb_get_unmapped_area().
Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Cc: Adam Litke <agl@us.ibm.com> Cc: Andi Kleen <ak@suse.de> Cc: "David S. Miller" <davem@davemloft.net> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Shane Huang [Fri, 31 Aug 2007 06:56:38 +0000 (23:56 -0700)]
i2c-piix4: Fix SB700 PCI device ID
We find that SB700 and SB800 use the same SMBus device ID as SB600, which is
0x4385, instead of the already submitted 0x4395.
Besides removing the wrong SB700 device ID, add SB800 support to kernel, by
renaming the PCI_DEVICE_ID_ATI_IXP600_SMBUS into
PCI_DEVICE_ID_ATI_SBX00_SMBUS.
Signed-off-by: Shane Huang <shane.huang@amd.com> Signed-off-by: Jean Delvare <khali@linux-fr.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Oleg Nesterov [Fri, 31 Aug 2007 06:56:35 +0000 (23:56 -0700)]
sigqueue_free: fix the race with collect_signal()
Spotted by taoyue <yue.tao@windriver.com> and Jeremy Katz <jeremy.katz@windriver.com>.
collect_signal: sigqueue_free:
list_del_init(&first->list);
if (!list_empty(&q->list)) {
// not taken
}
q->flags &= ~SIGQUEUE_PREALLOC;
__sigqueue_free(first); __sigqueue_free(q);
Now, __sigqueue_free() is called twice on the same "struct sigqueue" with the
obviously bad implications.
In particular, this double free breaks the array_cache->avail logic, so the
same sigqueue could be "allocated" twice, and the bug can manifest itself via
the "impossible" BUG_ON(!SIGQUEUE_PREALLOC) in sigqueue_free/send_sigqueue.
Hopefully this can explain these mysterious bug-reports, see
Alexey Dobriyan reports this patch makes the difference for the testcase, but
nobody has an access to the application which opened the problems originally.
Also, this patch removes tasklist lock/unlock, ->siglock is enough.
This shows that a driver's irq handler was running both in hard interrupt
and process contexts with irqs enabled. The latter was done during
free_irq() call and was possible only with CONFIG_DEBUG_SHIRQ enabled.
This was fixed by another patch.
But similar problem is possible with request_irq(): any locks taken from
irq handler could be vulnerable - especially with soft interrupts. This
patch fixes it by disabling local interrupts during handler's run. (It
seems, disabling softirqs should be enough, but it needs more checking
on possible races or other special cases).
Ryusuke Konishi [Fri, 31 Aug 2007 06:56:33 +0000 (23:56 -0700)]
eCryptfs: fix possible fault in ecryptfs_sync_page
This will avoid a possible fault in ecryptfs_sync_page().
In the function, eCryptfs calls sync_page() method of a lower filesystem
without checking its existence. However, there are many filesystems that
don't have this method including network filesystems such as NFS, AFS, and
so forth. They may fail when an eCryptfs page is waiting for lock.
Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp> Acked-by: Michael Halcrow <mhalcrow@us.ibm.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Andrew Victor [Fri, 31 Aug 2007 06:56:32 +0000 (23:56 -0700)]
Fix typo in atmel_spi.c
Fix cut 'n paste bug in Atmel SPI driver.
Signed-off-by: Andrew Victor <andrew@sanpeople.com> Acked-by: David Brownell <david-b@pacbell.net> Acked-by: Haavard Skinnemoen <hskinnemoen@atmel.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
projects / linux-2.6 / log