Subject: lockdep: annotate do_debug() trap handler Ensure the hardirq state is consistent before using locks. Use the rare trace_hardirqs_fixup() because the trap can happen in any context. Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl> --- arch/x86/kernel/traps_32.c | 2 ++ arch/x86/kernel/traps_64.c | 2 ++ 2 files changed, 4 insertions(+) Index: linux-2.6/arch/x86/kernel/traps_32.c =================================================================== --- linux-2.6.orig/...

I've just released Linux 2.4.35.4. It fixes very minor issues, but some patches have been resting here for a long time, and there was no reason to hold them. Two minor old vulnerabilities were fixed : - CVE-2006-5823 would cause the kernel to oops on specially crafted CRAMFS filesystems, though on 2.4.35 we never got more than an error in the logs. The fix was merged anyway. - CVE-2004-2731 corresponds to an incorrect size check in sparc's openprom driver. It was fixed in 2.5 but not in...

Without rcu/tasklist/siglock lock task_pid_nr_ns() may read the freed memory, move the callsite under ->siglock. Sadly, we can report pid == 0 if the task was detached. Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru> --- 24/fs/proc/array.c~dtst 2007-11-09 12:57:30.000000000 +0300 +++ 24/fs/proc/array.c 2007-11-17 21:26:55.000000000 +0300 @@ -392,7 +392,7 @@ static int do_task_stat(struct task_stru sigset_t sigign, sigcatch; char state; int res; - pid_t ppid = 0, pgid = -1, sid ...

I upgraded today from 2.6.23 to 2.6.23.8 and started seeing a lot of these in the logs: BUG: soft lockup detected on CPU#0! [<c011d761>] update_process_times+0x32/0x54 [<c012b06f>] tick_sched_timer+0x5e/0x99 [<c01274db>] hrtimer_interrupt+0x112/0x197 [<c012b011>] tick_sched_timer+0x0/0x99 [<c010f02c>] smp_apic_timer_interrupt+0x60/0x6f [<c01cd9e7>] acpi_hw_register_write+0x118/0x148 [<c0104620>] apic_timer_interrupt+0x28/0x30 [<f8827350>] a...

This is the culprit, reverting fixes the issue. Cheers David --- a/kernel/softlockup.c +++ b/kernel/softlockup.c @@ -80,10 +80,11 @@ void softlockup_tick(void) print_timestamp = per_cpu(print_timestamp, this_cpu); /* report at most once a second */ - if (print_timestamp < (touch_timestamp + 1) || - did_panic || - !per_cpu(watchdog_task, this_cpu)) + if ((print_timestamp >= touch_timestamp && + ...

Confirmed on my server machine. Seems to coincide with cpufreq processor speed changes. Config available on request, (system is also a single core athlon) Cheers -

- Make pci_slot the primary sysfs entity. hotplug_slot becomes a subsidiary structure. o pci_create_slot() creates and registers a slot with the PCI core o pci_slot_add_hotplug() gives it hotplug capability - Change the prototype of pci_hp_register() to take the bus and slot number (on parent bus) as parameters. - Remove all the ->get_address methods since this functionality is now handled by pci_slot directly. v2 -> v3: Separated slot creation and slot hotplu...

Register one slot per slot, rather than one slot per function. Change the name of the slot to fake%d instead of the pci address. Signed-off-by: Alex Chiang <achiang@hp.com> Signed-off-by: Matthew Wilcox <matthew@wil.cx> --- drivers/pci/hotplug/fakephp.c | 80 +++++++++++++++------------------------- 1 files changed, 30 insertions(+), 50 deletions(-) diff --git a/drivers/pci/hotplug/fakephp.c b/drivers/pci/hotplug/fakephp.c index 027f686..996c942 100644 --- a/drivers/pci/hotplug/f...

Hi, 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() right now lacks 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 de...

With this change we will see zillions of files being rebuild each time we pick up another kernel version from git and friends. For me it looks like this right now: #define UTS_RELEASE "2.6.24-rc2-g99fee6d7-dirty" committing my local changes made it look like: #define UTS_RELEASE "2.6.24-rc2-g99fee6d7" The above change will trigger a rebuild of all files that reference UTS_RELEASE as will all WARN_ON users. And this is with the default configuration. So if we want this then we want to pu...

On Sat, 17 Nov 2007 10:27:20 -0800 by ... not too much at least, gcc ought to be quite good at merging same-strings into one, so it's just one extra pointer argument -- If you want to reach me at my work email, use arjan@linux.intel.com For development, discussion and tips for power savings, visit http://www.lesswatts.org -

On Sat, 17 Nov 2007 10:46:52 -0800 ok so how about putting the same into dump_stack() instead? (see below) added bonus is that it's now present for all dumps that use dump_stack(), not just WARN_ON() (the format I copied from the exact line used by oopses) Subject: printk kernel version in WARN_ON and other dump_stack users From: Arjan van de Ven <arjan@linux.intel.com> today, all oopses contain a version number of the kernel, which is nice because the people who actually do bother t...

proc_pid_readdir: for (...; ...; task = next_tgid(tgid + 1, ns)) { tgid = task_pid_nr_ns(task, ns); ... use tgid ... The first problem is that task_pid_nr_ns() can race with RCU and read the freed memory. However, rcu_read_lock() can't help. next_tgid() returns a pinned task_struct, but the task can be released (and it's pid detached) before task_pid_nr_ns() reads the pid_t value. In that case task_pid_nr_ns() returns 0 thus breaking the whole logic. Make sure that task_pid_nr_ns() re...

I would like to request a new boot flag/kernel option that would make the following scenario possible: 1) Working on laptop with a live USB distro on a read-only USB stick. 2) Suspend laptop. 3) Detach USB stick. 4) Do other things, get on a plane, go on a bus, deal with police officer giving you a ticket for operating a laptop while driving... 5) Attach the same read-only USB stick. 6) Resume laptop. 7) Continue work as if nothing happened. The last time we were able to do something like ...

Add support to wireless usb key sitecom wl162 in ./drivers/net/wireless/zd1211rw/zd_usb.c added srtuct { USB_DEVICE(0x0cde, 0x001a), .driver_info = DEVICE_ZD1211B }, I've to do it by myself every time i update kernel. Please add it. Thank you -

Kai Ruhnau <kai@tragetaschen.dyndns.org> : No idea but it seems to be plaguing us: - sky2 Kai Ruhnau <kai@tragetaschen.dyndns.org> - r8169 + Abit fatal1ty motherboard (so far nobody reported a 8169 on a fatal1ty motherboard with a sensible vendor id) Josh Logan <joshtlogan@gmail.com> Ciaran McCreesh <ciaran.mccreesh@blueyonder.co.uk> -- Ueimor -

FYI, Just added Sam's latest fix for the x86 build mechanism on top of the patches below. Thanks, -

.. btw, Thomas, you forgot to add your sign-off to your version. Linus -

Simplify "make ARCH=x86" and fix kconfig so we again can set 64BIT in all.config. For a fix the diffstat is nice: 6 files changed, 3 insertions(+), 36 deletions(-) The patch reverts these commits: 0f855aa64b3f63d35a891510cf7db932a435c116 -> kconfig: add helper to set config symbol from environment variable 2a113281f5cd2febbab21a93c8943f8d3eece4d3 -> kconfig: use $K64BIT to set 64BIT with all*config targets Roman Zippel pointed out that kconfig supported string compares so the additi...

If you are accessing a scratched Video DVD and the device cannot read it, the process ends. What about a more tolerant way to handle unreadable blocks. Especially on Video DVDs single blocks are not that important than on data dvds. So is there a way that the kernel tells the device to skip these bad blocks? Tobias -

Fix fallout from the x86 32/64-bit merge. asm-x86/setup_32.h no longer exists. Signed-off-by: Peter Hagervall <hager@cs.umu.se> --- diff --git a/include/asm-x86/mach-voyager/setup_arch.h b/include/asm-x86/mach-voyager/setup_arch.h index 1710ae1..71729ca 100644 --- a/include/asm-x86/mach-voyager/setup_arch.h +++ b/include/asm-x86/mach-voyager/setup_arch.h @@ -1,5 +1,5 @@ #include <asm/voyager.h> -#include <asm/setup_32.h> +#include <asm/setup.h> #define VOYAGER_BIO...

Use kasprintf instead of kmalloc()-strcpy()-strcat(). Cc: Anton Vorontsov <cbou@mail.ru> Cc: David Woodhouse <dwmw2@infradead.org> Signed-off-by: Akinobu Mita <akinobu.mita@gmail.com> --- drivers/power/power_supply_leds.c | 25 +++++++------------------ 1 file changed, 7 insertions(+), 18 deletions(-) Index: 2.6-mm/drivers/power/power_supply_leds.c =================================================================== --- 2.6-mm.orig/drivers/power/power_supply_leds.c +++ 2....

Please add this information to the bug entry at http://bugzilla.kernel.org/show_bug.cgi?id=9346 Thanks, Rafael -

>I would also like to see something like this so if _you_ have common sense, it`s up to you alone to make your own life easier and you should try hard to change their mind. how can you make sure your code is good and stable if other`s don`t review it ? why should you constantly work on porting your code to the next kernel release ? even worse, you need to port it to different vendor specific kernels. that`s a real waste of time... please try changing their minds and if you need some help/arguments...

This patch adds the algorithm to pull tasks from RT overloaded runqueues. When a pull RT is initiated, all overloaded runqueues are examined for a RT task that is higher in prio than the highest prio task queued on the target runqueue. If another runqueue holds a RT task that is of higher prio than the highest prio task on the target runqueue is found it is pulled to the target runqueue. Signed-off-by: Steven Rostedt <srostedt@redhat.com> --- kernel/sched.c | 2 kernel/sched_rt....