lkml.org 
[lkml]   [2008]   [Jun]   [11]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
Patch in this message
/
SubjectRe: >3G => iommu => suspend problems -- was Re: SB600 AHCI: Hard Disk Corruption
From
Date
Hello

On Mon, 2008-06-09 at 00:36 +0200, Rafael J. Wysocki wrote:

> > > It's on top of the patch that adds the GART sysdev.

> Updated patch follows. It has been tested a little on my new 4 GB test box on
> which 2.6.26-rc4 failed miserably with severe consequences. More testing
> welcome, but please be careful.

No risk no fun! :-)

As the suspend-vs-iommu-prevent-suspend-if-we-could-not-resume.patch,
where the GART sysdev is added, is now included in 2.6.26-rc5, I was
able to apply this one seamlessly to it (-rc5 tree) and have just
rebooted *, suspended to disk and ram once wiht all filesystems (>400GB)
mounted rw ** and everything is _working fine_.

As usual, a kernel log, the applied patch and a test report are
available at http://zefir.890m.com/kernel-testing/ .

pat@babar:~/tmp/dmesg$ grep -C3 GART dmesg.2.6.26-rc5-gart-suspend.txt
[ 0.294419] PCI: Using ACPI for IRQ routing
[ 0.314420] PCI-DMA: Disabling AGP.
[ 0.314420] PCI-DMA: aperture base @ 4000000 size 65536 KB
[ 0.314420] PCI-DMA: using GART IOMMU.
[ 0.314420] PCI-DMA: Reserving 64MB of IOMMU area in the AGP aperture
[ 0.314420] hpet0: at MMIO 0xfed00000, IRQs 2, 8, 0, 0
[ 0.314420] hpet0: 4 32-bit timers, 14318180 Hz
--
[ 221.770810] CPU1 is down
[ 221.770966] PM: Creating hibernation image:
[ 221.772014] PM: Need to copy 220260 pages
[ 221.772014] PCI-DMA: Resuming GART IOMMU
[ 221.772014] Enabling non-boot CPUs ...
[ 221.772014] CPU0 attaching NULL sched-domain.
[ 221.781073] SMP alternatives: switching to SMP code
--
[ 251.756678] groups: 0
[ 251.756984] CPU1 is down
[ 251.756984] Back to C!
[ 251.756988] PCI-DMA: Resuming GART IOMMU
[ 251.757355] Enabling non-boot CPUs ...
[ 251.757515] CPU0 attaching NULL sched-domain.
[ 251.766920] SMP alternatives: switching to SMP code



* Rebooted for the first time since i compiled and rebooted -rc4 with
a similar "home-grown" patch on 31. of May, with which I suspended
several times to disk and ram without any problems.

** and also generally the same setup where suspending has disastrous
effects using kernel < 2.6.26-rc5

> ---
> Add resume handling to GART IOMMU.
>
> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
> ---
> arch/x86/kernel/aperture_64.c | 2 +
> arch/x86/kernel/pci-gart_64.c | 75 +++++++++++++++++++++++++++++++++---------
> include/asm-x86/gart.h | 1
> 3 files changed, 62 insertions(+), 16 deletions(-)
>
> Index: linux-2.6/arch/x86/kernel/pci-gart_64.c
> ===================================================================
> --- linux-2.6.orig/arch/x86/kernel/pci-gart_64.c
> +++ linux-2.6/arch/x86/kernel/pci-gart_64.c
> @@ -549,14 +549,70 @@ static __init unsigned read_aperture(str
> return aper_base;
> }
>
> +static void enable_gart_translations(void)
> +{
> + int i;
> +
> + for (i = 0; i < num_k8_northbridges; i++) {
> + struct pci_dev *dev;
> + u32 gatt_reg;
> + u32 ctl;
> +
> + dev = k8_northbridges[i];
> + gatt_reg = __pa(agp_gatt_table) >> 12;
> + gatt_reg <<= 4;
> + pci_write_config_dword(dev, 0x98, gatt_reg);
> + pci_read_config_dword(dev, 0x90, &ctl);
> +
> + ctl |= 1;
> + ctl &= ~((1<<4) | (1<<5));
> +
> + pci_write_config_dword(dev, 0x90, ctl);
> + }
> +}
> +
> +/*
> + * If fix_up_north_bridges is set, the north bridges have to be fixed up on
> + * resume in the same way as they are handled in gart_iommu_hole_init().
> + */
> +static bool fix_up_north_bridges;
> +static u32 aperture_order;
> +static u32 aperture_alloc;
> +
> +void set_up_gart_resume(u32 aper_order, u32 aper_alloc)
> +{
> + fix_up_north_bridges = true;
> + aperture_order = aper_order;
> + aperture_alloc = aper_alloc;
> +}
> +
> static int gart_resume(struct sys_device *dev)
> {
> + printk(KERN_INFO "PCI-DMA: Resuming GART IOMMU\n");
> +
> + if (fix_up_north_bridges) {
> + int i;
> +
> + for (i = 0; i < num_k8_northbridges; i++) {
> + struct pci_dev *dev = k8_northbridges[i];
> +
> + /*
> + * Don't enable translations just yet. That is the next
> + * step. Restore the pre-suspend aperture settings.
> + */
> + pci_write_config_dword(dev, 0x90, aperture_order << 1);
> + pci_write_config_dword(dev, 0x94, aperture_alloc >> 25);
> + }
> + }
> +
> + enable_gart_translations();
> +
> return 0;
> }
>
> static int gart_suspend(struct sys_device *dev, pm_message_t state)
> {
> - return -EINVAL;
> + return 0;
> }
>
> static struct sysdev_class gart_sysdev_class = {
> @@ -614,27 +670,14 @@ static __init int init_k8_gatt(struct ag
> memset(gatt, 0, gatt_size);
> agp_gatt_table = gatt;
>
> - for (i = 0; i < num_k8_northbridges; i++) {
> - u32 gatt_reg;
> - u32 ctl;
> -
> - dev = k8_northbridges[i];
> - gatt_reg = __pa(gatt) >> 12;
> - gatt_reg <<= 4;
> - pci_write_config_dword(dev, 0x98, gatt_reg);
> - pci_read_config_dword(dev, 0x90, &ctl);
> -
> - ctl |= 1;
> - ctl &= ~((1<<4) | (1<<5));
> -
> - pci_write_config_dword(dev, 0x90, ctl);
> - }
> + enable_gart_translations();
>
> error = sysdev_class_register(&gart_sysdev_class);
> if (!error)
> error = sysdev_register(&device_gart);
> if (error)
> panic("Could not register gart_sysdev -- would corrupt data on next suspend");
> +
> flush_gart();
>
> printk(KERN_INFO "PCI-DMA: aperture base @ %x size %u KB\n",
> Index: linux-2.6/arch/x86/kernel/aperture_64.c
> ===================================================================
> --- linux-2.6.orig/arch/x86/kernel/aperture_64.c
> +++ linux-2.6/arch/x86/kernel/aperture_64.c
> @@ -413,4 +413,6 @@ void __init gart_iommu_hole_init(void)
> write_pci_config(0, num, 3, 0x90, aper_order<<1);
> write_pci_config(0, num, 3, 0x94, aper_alloc>>25);
> }
> +
> + set_up_gart_resume(aper_order, aper_alloc);
> }
> Index: linux-2.6/include/asm-x86/gart.h
> ===================================================================
> --- linux-2.6.orig/include/asm-x86/gart.h
> +++ linux-2.6/include/asm-x86/gart.h
> @@ -11,6 +11,7 @@ extern void gart_iommu_shutdown(void);
> extern void __init gart_parse_options(char *);
> extern void early_gart_iommu_check(void);
> extern void gart_iommu_hole_init(void);
> +extern void set_up_gart_resume(u32, u32);
> extern int fallback_aper_order;
> extern int fallback_aper_force;
> extern int gart_iommu_aperture;

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/

\
 
 \ /
  Last update: 2008-06-11 13:45    [W:0.552 / U:0.188 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site