24 Jan, 2007
16 commits
-
Any newly added irq handler may obviously make any old spurious irq
status invalid, since the new handler may well be the thing that is
supposed to handle any interrupts that came in.So just clear the statistics when adding handlers.
Pointed-out-by: Alan Cox
Acked-by: Thomas Gleixner
Acked-by: Ingo Molnar
Signed-off-by: Linus Torvalds -
* 'upstream' of git://ftp.linux-mips.org/pub/scm/upstream-linus:
[MIPS] Vr41xx: Fix after GENERIC_HARDIRQS_NO__DO_IRQ change
[MIPS] SMTC: Instant IPI replay. -
As written, sys_shmget will return ENOSPC when one page is still
available for allocation. This patch corrects the test.Signed-off-by: Guy Streeter
Signed-off-by: Linus Torvalds
-- -
* 'for-linus' of master.kernel.org:/pub/scm/linux/kernel/git/roland/infiniband:
IB/ehca: Fix mismatched spin_unlock in irq handler
IB/ehca: Fix improper use of yield() with spinlock held
IB/srp: Check match_strdup() return -
Signed-off-by: Al Viro
Signed-off-by: Linus Torvalds -
memset() after kmalloc() on size * 8 would better be on size * 8, not
just size; fixed by switching to kcalloc() - it's more idiomatic anyway.Signed-off-by: Al Viro
Signed-off-by: Linus Torvalds -
Signed-off-by: Al Viro
Signed-off-by: Linus Torvalds -
Signed-off-by: Al Viro
Signed-off-by: Linus Torvalds -
while lock-profiling the -rt kernel i noticed weird contention during
mmap-intense workloads, and the tracer showed the following gem, in one
of our MM hotpaths:threaded-2771 1.... 65us : sys_munmap (sysenter_do_call)
threaded-2771 1.... 66us : profile_munmap (sys_munmap)
threaded-2771 1.... 66us : blocking_notifier_call_chain (profile_munmap)
threaded-2771 1.... 66us : rt_down_read (blocking_notifier_call_chain)ouch! a global rw-semaphore taken in one of the most performance-
sensitive codepaths of the kernel. And i dont even have oprofile
enabled! All distro kernels have CONFIG_PROFILING enabled, so this
scalability problem affects the majority of Linux users.The fix is to enhance blocking_notifier_call_chain() to only take the
lock if there appears to be work on the call-chain.With this patch applied i get nicely saturated system, and much higher
munmap performance, on SMP systems.And as a bonus this also fixes a similar scalability bottleneck in the
thread-exit codepath: profile_task_exit() ...Signed-off-by: Ingo Molnar
Acked-by: Peter Zijlstra
Acked-by: Nick Piggin
Signed-off-by: Linus Torvalds -
* 'linus' of master.kernel.org:/pub/scm/linux/kernel/git/perex/alsa:
[ALSA] Repair snd-usb-usx2y over OHCI -
* 'upstream-linus' of master.kernel.org:/pub/scm/linux/kernel/git/jgarzik/netdev-2.6:
NetXen: Use pci_register_driver() instead of pci_module_init() in init_module
NetXen: Firmware check modifications
ehea: Fixed possible nullpointer access
ehea: Added logging off associated errors
ehea: Improved logging of permission issues
ehea: New method to determine number of available ports
ehea: Modified initial autoneg state determination
ehea: Fixing firmware queue config issue
ehea: Fixed wrong dereferencation
PHY: Export phy ethtool helpers
modify 3c589_cs to be SMP safe -
* 'ftape' of master.kernel.org:/pub/scm/linux/kernel/git/jgarzik/misc-2.6:
more ftape removal -
* 'kill-jffs-prep' of master.kernel.org:/pub/scm/linux/kernel/git/jgarzik/misc-2.6:
Note that JFFS (v1) is to be deleted, in feature-removal-schedule.txt -
A flag was recently added to the elevator code to avoid
performing an unplug when reuests are being re-queued.
The goal of this flag was to avoid a deep recursion that
can occur when re-queueing requests after a SCSI device/host
reset. See http://lkml.org/lkml/2006/5/17/254However, that fix added the flag near the bottom of a case
statement, where an earlier break (in an if statement) could
transport one out of the case, without setting the flag.
This patch sets the flag earlier in the case statement.I re-discovered the deep recursion recently during testing;
I was told that it was a known problem, and the fix to it was
in the kernel I was testing. Indeed it was ... but it didn't
fix the bug. With the patch below, I no longer see the bug.Signed-off by: Linas Vepstas
Signed-off-by: Jens Axboe
Cc: Chris Wright
Signed-off-by: Linus Torvalds -
Signed-off-by: Yoichi Yuasa
Signed-off-by: Ralf Baechle -
SMTC pseudo-interrupts between TCs are deferred and queued if the target
TC is interrupt-inhibited (IXMT). In the first SMTC prototypes, these
queued IPIs were serviced on return to user mode, or on entry into the
kernel idle loop. The INSTANT_REPLAY option dispatches them as part of
local_irq_restore() processing, which adds runtime overhead (hence the
option to turn it off), but ensures that IPIs are handled promptly even
under heavy I/O interrupt load.Signed-off-by: Ralf Baechle
23 Jan, 2007
24 commits
-
Seems to be some left-over debug code.
Cc: Len Brown
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
this patch fills in the portions for ia64 kexec.
Signed-off-by: Simon Horman
Cc: "Zou, Nanhai"
Cc: Vivek Goyal
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
Mohan Kumar suggested making kexec-tools-testing.tar.gz a link to the
latest version. I have done this and this patch updates the documentation
accordingly.Signed-off-by: Simon Horman
Cc: Vivek Goyal
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
This patch fixes a confusion reiserfs has for a long time.
On release file operation reiserfs used to try to pack file data stored in
last incomplete page of some files into metadata blocks. After packing the
page got cleared with clear_page_dirty. It did not take into account that
the page may be mmaped into other process's address space. Recent
replacement for clear_page_dirty cancel_dirty_page found the confusion with
sanity check that page has to be not mapped.The patch fixes the confusion by making reiserfs avoid tail packing if an
inode was ever mmapped. reiserfs_mmap and reiserfs_file_release are
serialized with mutex in reiserfs specific inode. reiserfs_mmap locks the
mutex and sets a bit in reiserfs specific inode flags.
reiserfs_file_release checks the bit having the mutex locked. If bit is
set - tail packing is avoided. This eliminates a possibility that mmapped
page gets cancel_page_dirty-ed.Signed-off-by: Vladimir Saveliev
Cc: Jeff Mahoney
Cc: Chris Mason
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
Currently one can specify an arbitrary node mask to mbind that includes
nodes not allowed. If that is done with an interleave policy then we will
go around all the nodes. Those outside of the currently allowed cpuset
will be redirected to the border nodes. Interleave will then create
imbalances at the borders of the cpuset.This patch restricts the nodes to the currently allowed cpuset.
The RFC for this patch was discussed at
http://marc.theaimsgroup.com/?t=116793842100004&r=1&w=2Signed-off-by: Christoph Lameter
Cc: Paul Jackson
Cc: Andi Kleen
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
The following patch fixes a few problems with the tlclk driver.
* bug in the select_amcb1_transmit_clock
* racy read sys call
* racy open sys call
* use of add_timer where mod_timer would be better
* change to the timer data parameter useSigned-off-by: Mark Gross
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
For large size DIO that needs multiple bio, one full page worth of data was
lost at the boundary of bio's maximum sector or segment limits. After a
bio is full and got submitted. The outer while (nbytes) { ... } loop will
allocate a new bio and just march on to index into next page. It just
forgets about the page that bio_add_page() rejected when previous bio is
full. Fix it by put the rejected page back to pvec so we pick it up again
for the next bio.Signed-off-by: Ken Chen
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
This fixes the SH rtc driver correctly act on the "enabled" flag when
setting an alarm.Signed-off-by: Jamie Lenehan
Cc: David Brownell
Cc: Alessandro Zummo
Cc: Paul Mundt
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
If a page is marked as dirty in the guest pte, set_pte_common() can set the
writable bit on newly-instantiated shadow pte. This optimization avoids
a write fault after the initial read fault.However, if a write fault instantiates the pte, fix_write_pf() incorrectly
reports the fault as a guest page fault, and the guest oopses on what appears
to be a correctly-mapped page.Fix is to detect the condition and only report a guest page fault on a user
access to a kernel page.With the fix, a kvm guest can survive a whole night of running the kernel
hacker's screensaver (make -j9 in a loop).Signed-off-by: Avi Kivity
Cc: Ingo Molnar
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
The various bit string instructions (bts, btc, etc.) fail to adjust the
address correctly if the bit address is beyond BITS_PER_LONG.This bug creeped in as the emulator originally relied on cr2 to contain the
memory address; however we now decode it from the mod r/m bits, and must
adjust the offset to account for large bit indices.The patch is rather large because it switches src and dst decoding around, so
that the bit index is available when decoding the memory address.This fixes workloads like the FC5 installer.
Signed-off-by: Avi Kivity
Cc: Ingo Molnar
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
The kvm mmio read path looks like:
1. guest read faults
2. kvm emulates read, calls emulator_read_emulated()
3. fails as a read requires userspace help
4. exit to userspace
5. userspace emulates read, kvm sets vcpu->mmio_read_completed
6. re-enter guest, fault again
7. kvm emulates read, calls emulator_read_emulated()
8. succeeds as vcpu->mmio_read_emulated is set
9. instruction completes and guest is resumedA problem surfaces if the userspace exit (step 5) also requests an interrupt
injection. In that case, the guest does not re-execute the original
instruction, but the interrupt handler. The next time an mmio read is
exectued (likely for a different address), step 3 will find
vcpu->mmio_read_completed set and return the value read for the original
instruction.The problem manifested itself in a few annoying ways:
- little squares appear randomly on console when switching virtual terminals
- ne2000 fails under nfs read load
- rtl8139 complains about "pci errors" even though the device model is
incapable of issuing them.Fix by skipping interrupt injection if an mmio read is pending.
A better fix is to avoid re-entry into the guest, and re-emulating immediately
instead. However that's a bit more complex.Signed-off-by: Avi Kivity
Cc: Ingo Molnar
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
This makes the vmwrite errors on vm shutdown go away.
Signed-off-by: Avi Kivity
Cc: Ingo Molnar
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
The paravirt subsystem is still in flux so all exports from it are
definitely internal use only. The APIs around this /will/ change.Signed-off-by: Ingo Molnar
Cc: Andi Kleen
Cc: Zachary Amsden
Cc: Jeremy Fitzhardinge
Acked-by: Rusty Russell
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
Sing the praises of `gcc -W'. Would have prevented that blockdev direct-IO
bug.Cc: "Chen, Kenneth W"
Cc: "Randy.Dunlap"
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
size_t is unsigned. IO errors aren't getting through.
Cc: "Chen, Kenneth W"
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
Commit f2802e7f571c05f9a901b1f5bd144aa730ccc88e and its x86 version
(b7471c6da94d30d3deadc55986cc38d1ff57f9ca) adds nmi_known_cpu() check
while parsing boot options in x86_64 and i386.With that, "nmi_watchdog=2" stops working for me on Intel Core 2 CPU
based system.The problem is, setup_nmi_watchdog is called while parsing the boot
option and identify_cpu is not done yet. So, the return value of
nmi_known_cpu() is not valid at this point.So revert that check. This should not have any adverse effect as the
nmi_known_cpu() check is done again later in enable_lapic_nmi_watchdog().Signed-off-by: Venkatesh Pallipadi
Cc: Don Zickus
Cc: Andi Kleen
Cc: Ingo Molnar
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
export profile_hits() on !SMP too.
Cc: Ingo Molnar
Cc: Avi Kivity
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
The previous patch 'Repair snd-usb-usx2y for usb 2.6.18' assumed
urb->start_frame roll over beyond MAX_INT for both UHCI & OHCI.
This isn't true until now (kernel 2.6.20).
Fix this by only looking at the common between OHCI & UHCI Frame number
range.
This is for mainline and stable kernels >= 2.6.18.Signed-off-by: Karsten Wiese
Signed-off-by: Takashi Iwai
Signed-off-by: Jaroslav Kysela -
This will use pci_register_driver() instead of pci_module_init().
Signed-off-by: Amit S. Kale
Signed-off-by: Richard Knutsson
Signed-off-by: Jeff Garzik -
This patch is to make the driver work with multiple minor firmware versions
Signed-off-by: Amit S. Kale
Signed-off-by: Jeff Garzik -
Fixed possible nullpointer access in event queue processing
Signed-off-by: Thomas Klein
Signed-off-by: Jeff Garzik -
Added logging of error events associated with a specific queue pair
Signed-off-by: Thomas Klein
Signed-off-by: Jeff Garzik -
Disabled dump of hcall regs on some permission issues and
fixed appropriate misleading logmessagesSigned-off-by: Thomas Klein
Signed-off-by: Jeff Garzik -
Count OFDT nodes to determine the number of available ports
instead of using the possibly outdated value from the hypervisorSigned-off-by: Thomas Klein
Signed-off-by: Jeff Garzik
