The patch titled genirq: do not disable IRQ_WAKEUP marked irqs on suspend has been added to the -mm tree. Its filename is genirq-do-not-disable-irq_wakeup-marked-irqs-on-suspend.patch Before you just go and hit "reply", please: a) Consider who else should be cc'ed b) Prefer to cc a suitable mailing list as well c) Ideally: find the original patch on the mailing list and do a reply-to-all to that, adding suitable additional cc's *** Remember to use Documentation/SubmitChecklist when testing your code *** See http://userweb.kernel.org/~akpm/stuff/added-to-mm.txt to find out what to do about this The current -mm tree may be found at http://userweb.kernel.org/~akpm/mmotm/ ------------------------------------------------------ Subject: genirq: do not disable IRQ_WAKEUP marked irqs on suspend From: Thomas Gleixner <tglx@xxxxxxxxxxxxx> commit 0a0c5168df (PM: Introduce functions for suspending and resuming device interrupts) iterates through all interrupts and disables them on the hardware level. Some architectures have functionality implemented to mark an interrupt source as wakeup source for suspend, but the new power management code disables them unconditionally which breaks the resume on interrupt functionality. The wakeup interrupts are marked in the status with the IRQ_WAKEUP bit. Skip the disablement for those interrupts which have the IRQ_WAKEUP bit set. rmk didst say: This sounds like a regression fix to me - if IRQs are disabled on the controllers, it means we have systems which can only ever be suspended. The only way to wake them back up would be to hard-reset them. IOW, the regression turns suspend into a hang. rmk didst further say: ARM systems can only be woken up by unmasked interrupts. If the current kernel masks all interrupts, suspending ARM systems is a _one_ way ticket to a full reboot. With all interrupts masked, there is NO WAY to wake up ARM interrupts. So I'm afraid my comment here without any further suggestion that the problem has been fixed in some other way is that this patch _is_ required _if_ the kernel is still masking all interrupts on suspend. Now, given that ARM systems were able to be resumed up until the offending commit was added to the kernel, it _is_ a regression, and I think under Linus' rules, a fix is not a fix if it causes regressions for some people. So IMHO it needs to be backed out and another approach taken, _if_ the problem is still present. Unfortunately, I've no time to verify if the problem is present at the moment. I'm sure someone else in the ARM community can do the necessary testing though. Eg, anyone with a PXA platform will do. Signed-off-by: Thomas Gleixner <tglx@xxxxxxxxxxxxx> Cc: "Rafael J. Wysocki" <rjw@xxxxxxx> Acked-by: Russell King <rmk@xxxxxxxxxxxxxxxx> Cc: Kevin Hilman <khilman@xxxxxxxxxxxxxxxxxxx> Cc: <stable@xxxxxxxxxx> Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx> --- kernel/irq/pm.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff -puN kernel/irq/pm.c~genirq-do-not-disable-irq_wakeup-marked-irqs-on-suspend kernel/irq/pm.c --- a/kernel/irq/pm.c~genirq-do-not-disable-irq_wakeup-marked-irqs-on-suspend +++ a/kernel/irq/pm.c @@ -29,7 +29,8 @@ void suspend_device_irqs(void) unsigned long flags; spin_lock_irqsave(&desc->lock, flags); - __disable_irq(desc, irq, true); + if (!(desc->status & IRQ_WAKEUP)) + __disable_irq(desc, irq, true); spin_unlock_irqrestore(&desc->lock, flags); } _ Patches currently in -mm which might be from tglx@xxxxxxxxxxxxx are origin.patch execve-must-clear-current-clear_child_tid-v2.patch lib-decompress_-only-include-linux-slabh-if-static-is-not-defined.patch linux-next.patch arch-x86-kernel-cpu-amdc-fix-x86_64-allnoconfig-build.patch kernel-core-add-smp_call_function_any.patch arch-x86-kernel-cpu-cpufreq-acpi-cpufreqc-avoid-cross-cpu-interrupts-by-using-smp_call_function_any.patch arch-x86-kernel-tscc-smi-workaround-for-pit_expect_msb.patch x86-add-reboot-fixup-for-compulab-sbc-fitpc2-board.patch ia32-use-generic-sys_pipe.patch preempt_active-add-default-defines.patch posix_cpu_timers_exit_group-do-not-use-thread_group_cputimer.patch timer-stats-fix-del_timer_sync-and-try_to_del_timer_sync.patch mtd-nand-add-page-parameter-to-all-read_page-read_page_raw-apis.patch mtd-nand-add-new-ecc-mode-ecc_hw_oob_first.patch mtd-nand-davinci-add-4-bit-ecc-support-for-large-page-nand-chips.patch genirq-do-not-disable-irq_wakeup-marked-irqs-on-suspend.patch time-add-function-to-convert-between-calendar-time-and-broken-down-time-for-universal-use.patch fatfs-use-common-time_to_tm-in-fat_time_unix2fat.patch linux-futexh-place-kernel-types-behind-__kernel__.patch futex-futex_wait_requeue_pi-commentary-corrections.patch include-linux-unaligned-lbe_byteshifth-fix-usage-for-compressed-kernels.patch lzo-add-support-for-lzo-compressed-kernels.patch lzo-arm-add-support-for-lzo-compressed-kernels.patch lzo-x86-add-support-for-lzo-compressed-kernels.patch lzo-add-lzo-compression-support-for-initramfs-and-old-style-initrd.patch -- To unsubscribe from this list: send the line "unsubscribe mm-commits" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html