From: "Rafael J. Wysocki" Subject: [Resend][PATCH -mm] Hibernation: Check if ACPI is enabled during restore in the right place Date: Wed, 12 Sep 2007 13:09:39 +0200 Message-ID: <200709121309.40687.rjw@sisk.pl> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-2 Content-Transfer-Encoding: QUOTED-PRINTABLE Return-path: Received: from ogre.sisk.pl ([217.79.144.158]:39035 "EHLO ogre.sisk.pl" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1760971AbXILK6H convert rfc822-to-8bit (ORCPT ); Wed, 12 Sep 2007 06:58:07 -0400 Content-Disposition: inline Sender: linux-acpi-owner@vger.kernel.org List-Id: linux-acpi@vger.kernel.org To: Andrew Morton Cc: ACPI Devel Maling List , Len Brown , LKML , Pavel Machek , pm list =46rom: Rafael J. Wysocki The following scenario leads to total confusion of the platform firmwar= e on some boxes (eg. HPC nx6325): * Hibernate with ACPI enabled * Resume passing "acpi=3Doff" to the boot kernel To prevent this from happening it's necessary to check if ACPI is enabl= ed (and enable it if that's not the case) _right_ _after_ control has been tran= sfered from the boot kernel to the image kernel, before device_power_up() is c= alled (ie. with interrupts disabled). =A0Enabling ACPI after calling device_p= ower_up() turns out to be insufficient. =46or this reason, introduce new hibernation callback ->leave() that wi= ll be executed before device_power_up() by the restored image kernel. =A0To m= ake it work, it also is necessary to move swsusp_suspend() from swsusp.c to di= sk.c (it's name is changed to "create_image", which is more up to the point)= =2E Signed-off-by: Rafael J. Wysocki Acked-by: Pavel Machek --- drivers/acpi/sleep/main.c | 7 ++++- include/linux/suspend.h | 7 +++++ kernel/power/disk.c | 58 +++++++++++++++++++++++++++++++++++++= ++++++++- kernel/power/power.h | 1=20 kernel/power/swsusp.c | 33 -------------------------- 5 files changed, 70 insertions(+), 36 deletions(-) Index: linux-2.6.23-rc6/drivers/acpi/sleep/main.c =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D --- linux-2.6.23-rc6.orig/drivers/acpi/sleep/main.c +++ linux-2.6.23-rc6/drivers/acpi/sleep/main.c @@ -238,13 +238,17 @@ static int acpi_hibernation_enter(void) return ACPI_SUCCESS(status) ? 0 : -EFAULT; } =20 -static void acpi_hibernation_finish(void) +static void acpi_hibernation_leave(void) { /* * If ACPI is not enabled by the BIOS and the boot kernel, we need to * enable it here. */ acpi_enable(); +} + +static void acpi_hibernation_finish(void) +{ acpi_leave_sleep_state(ACPI_STATE_S4); acpi_disable_wakeup_device(ACPI_STATE_S4); =20 @@ -274,6 +278,7 @@ static struct platform_hibernation_ops a .finish =3D acpi_hibernation_finish, .prepare =3D acpi_hibernation_prepare, .enter =3D acpi_hibernation_enter, + .leave =3D acpi_hibernation_leave, .pre_restore =3D acpi_hibernation_pre_restore, .restore_cleanup =3D acpi_hibernation_restore_cleanup, }; Index: linux-2.6.23-rc6/include/linux/suspend.h =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D --- linux-2.6.23-rc6.orig/include/linux/suspend.h +++ linux-2.6.23-rc6/include/linux/suspend.h @@ -156,6 +156,12 @@ extern void mark_free_pages(struct zone=20 * Called after the nonboot CPUs have been disabled and all of the low * level devices have been shut down (runs with IRQs off). * + * @leave: Perform the first stage of the cleanup after the system sle= ep state + * indicated by @set_target() has been left. + * Called right after the control has been passed from the boot kernel= to + * the image kernel, before the nonboot CPUs are enabled and before de= vices + * are resumed. Executed with interrupts disabled. + * * @pre_restore: Prepare system for the restoration from a hibernation= image. * Called right after devices have been frozen and before the nonboot * CPUs are disabled (runs with IRQs on). @@ -170,6 +176,7 @@ struct platform_hibernation_ops { void (*finish)(void); int (*prepare)(void); int (*enter)(void); + void (*leave)(void); int (*pre_restore)(void); void (*restore_cleanup)(void); }; Index: linux-2.6.23-rc6/kernel/power/disk.c =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D --- linux-2.6.23-rc6.orig/kernel/power/disk.c +++ linux-2.6.23-rc6/kernel/power/disk.c @@ -93,6 +93,17 @@ static int platform_pre_snapshot(int pla } =20 /** + * platform_leave - prepare the machine for switching to the normal mo= de + * of operation using the platform driver (called with interrupts disa= bled) + */ + +static void platform_leave(int platform_mode) +{ + if (platform_mode && hibernation_ops) + hibernation_ops->leave(); +} + +/** * platform_finish - switch the machine to the normal mode of operatio= n * using the platform driver (must be called after platform_prepare()) */ @@ -129,6 +140,51 @@ static void platform_restore_cleanup(int } =20 /** + * create_image - freeze devices that need to be frozen with interrupt= s + * off, create the hibernation image and thaw those devices. Control + * reappears in this routine after a restore. + */ + +int create_image(int platform_mode) +{ + int error; + + error =3D arch_prepare_suspend(); + if (error) + return error; + + local_irq_disable(); + /* At this point, device_suspend() has been called, but *not* + * device_power_down(). We *must* call device_power_down() now. + * Otherwise, drivers for some devices (e.g. interrupt controllers) + * become desynchronized with the actual state of the hardware + * at resume time, and evil weirdness ensues. + */ + error =3D device_power_down(PMSG_FREEZE); + if (error) { + printk(KERN_ERR "Some devices failed to power down, " + KERN_ERR "aborting suspend\n"); + goto Enable_irqs; + } + + save_processor_state(); + error =3D swsusp_arch_suspend(); + if (error) + printk(KERN_ERR "Error %d while creating the image\n", error); + /* Restore control flow magically appears here */ + restore_processor_state(); + if (!in_suspend) + platform_leave(platform_mode); + /* NOTE: device_power_up() is just a resume() for devices + * that suspended with irqs off ... no overall powerup. + */ + device_power_up(); + Enable_irqs: + local_irq_enable(); + return error; +} + +/** * hibernation_snapshot - quiesce devices and create the hibernation * snapshot image. * @platform_mode - if set, use the platform driver, if available, to @@ -163,7 +219,7 @@ int hibernation_snapshot(int platform_mo if (!error) { if (hibernation_mode !=3D HIBERNATION_TEST) { in_suspend =3D 1; - error =3D swsusp_suspend(); + error =3D create_image(platform_mode); /* Control returns here after successful restore */ } else { printk("swsusp debug: Waiting for 5 seconds.\n"); Index: linux-2.6.23-rc6/kernel/power/power.h =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D --- linux-2.6.23-rc6.orig/kernel/power/power.h +++ linux-2.6.23-rc6/kernel/power/power.h @@ -183,7 +183,6 @@ extern int swsusp_swap_in_use(void); extern int swsusp_check(void); extern int swsusp_shrink_memory(void); extern void swsusp_free(void); -extern int swsusp_suspend(void); extern int swsusp_resume(void); extern int swsusp_read(unsigned int *flags_p); extern int swsusp_write(unsigned int flags); Index: linux-2.6.23-rc6/kernel/power/swsusp.c =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D --- linux-2.6.23-rc6.orig/kernel/power/swsusp.c +++ linux-2.6.23-rc6/kernel/power/swsusp.c @@ -270,39 +270,6 @@ int swsusp_shrink_memory(void) return 0; } =20 -int swsusp_suspend(void) -{ - int error; - - if ((error =3D arch_prepare_suspend())) - return error; - - local_irq_disable(); - /* At this point, device_suspend() has been called, but *not* - * device_power_down(). We *must* device_power_down() now. - * Otherwise, drivers for some devices (e.g. interrupt controllers) - * become desynchronized with the actual state of the hardware - * at resume time, and evil weirdness ensues. - */ - if ((error =3D device_power_down(PMSG_FREEZE))) { - printk(KERN_ERR "Some devices failed to power down, aborting suspend= \n"); - goto Enable_irqs; - } - - save_processor_state(); - if ((error =3D swsusp_arch_suspend())) - printk(KERN_ERR "Error %d suspending\n", error); - /* Restore control flow magically appears here */ - restore_processor_state(); - /* NOTE: device_power_up() is just a resume() for devices - * that suspended with irqs off ... no overall powerup. - */ - device_power_up(); - Enable_irqs: - local_irq_enable(); - return error; -} - int swsusp_resume(void) { int error; - To unsubscribe from this list: send the line "unsubscribe linux-acpi" i= n the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html