Message ID | 201105120025.34531.rjw@sisk.pl (mailing list archive) |
---|---|
State | Not Applicable, archived |
Headers | show |
On Thu, 12 May 2011 00:25:34 +0200 "Rafael J. Wysocki" <rjw@sisk.pl> wrote: > There seems to be a problem on my Acer box which is related to DIPM > and switching the power source. Namely, when I detach the AC adapter > from the machine, the disk (which is an Intel SSD) freezes for a > while and something like this appears in dmesg: [...] Since updating to 2.6.39-rc7 (from 2.6.38.x) I have similar problems with two notebooks using Intel ICH: Acer PTZ1825 00:1f.2 SATA controller: Intel Corporation ICH9M/M-E SATA AHCI Controller (rev 03) (prog-if 01 [AHCI 1.0]) Subsystem: Acer Incorporated [ALI] Device 0300 Flags: bus master, 66MHz, medium devsel, latency 0, IRQ 42 I/O ports at 30e8 [size=8] I/O ports at 30fc [size=4] I/O ports at 30e0 [size=8] I/O ports at 30f8 [size=4] I/O ports at 3020 [size=32] Memory at d4504000 (32-bit, non-prefetchable) [size=2K] Capabilities: [80] MSI: Enable+ Count=1/16 Maskable- 64bit- Capabilities: [70] Power Management version 3 Capabilities: [a8] SATA HBA v1.0 Capabilities: [b0] PCI Advanced Features Kernel driver in use: ahci Dell Latitude E6510 00:1f.2 SATA controller: Intel Corporation 5 Series/3400 Series Chipset 6 port SATA AHCI Cont roller (rev 05) (prog-if 01 [AHCI 1.0]) Subsystem: Dell Device 040b Flags: bus master, 66MHz, medium devsel, latency 0, IRQ 41 I/O ports at 8090 [size=8] I/O ports at 8080 [size=4] I/O ports at 8070 [size=8] I/O ports at 8060 [size=4] I/O ports at 8020 [size=32] Memory at e9640000 (32-bit, non-prefetchable) [size=2K] Capabilities: [80] MSI: Enable+ Count=1/1 Maskable- 64bit- Capabilities: [70] Power Management version 3 Capabilities: [a8] SATA HBA v1.0 Capabilities: [b0] PCI Advanced Features Kernel driver in use: ahci Patch not tried yet (do not know / not looked into yet where to change for this chipset). Seems to be a 2.6.39-rcX regression?
On Friday, May 13, 2011, Michael Leun wrote: > On Thu, 12 May 2011 00:25:34 +0200 > "Rafael J. Wysocki" <rjw@sisk.pl> wrote: > > > There seems to be a problem on my Acer box which is related to DIPM > > and switching the power source. Namely, when I detach the AC adapter > > from the machine, the disk (which is an Intel SSD) freezes for a > > while and something like this appears in dmesg: > [...] > > Since updating to 2.6.39-rc7 (from 2.6.38.x) I have similar problems > with two notebooks using Intel ICH: > > Acer PTZ1825 > 00:1f.2 SATA controller: Intel Corporation ICH9M/M-E SATA AHCI > Controller (rev 03) (prog-if 01 [AHCI 1.0]) Subsystem: Acer > Incorporated [ALI] Device 0300 Flags: bus master, 66MHz, medium devsel, > latency 0, IRQ 42 I/O ports at 30e8 [size=8] > I/O ports at 30fc [size=4] > I/O ports at 30e0 [size=8] > I/O ports at 30f8 [size=4] > I/O ports at 3020 [size=32] > Memory at d4504000 (32-bit, non-prefetchable) [size=2K] > Capabilities: [80] MSI: Enable+ Count=1/16 Maskable- 64bit- > Capabilities: [70] Power Management version 3 > Capabilities: [a8] SATA HBA v1.0 > Capabilities: [b0] PCI Advanced Features > Kernel driver in use: ahci > > > Dell Latitude E6510 > 00:1f.2 SATA controller: Intel Corporation 5 Series/3400 Series Chipset > 6 port SATA AHCI Cont roller (rev 05) (prog-if 01 [AHCI 1.0]) > Subsystem: Dell Device 040b > Flags: bus master, 66MHz, medium devsel, latency 0, IRQ 41 > I/O ports at 8090 [size=8] > I/O ports at 8080 [size=4] > I/O ports at 8070 [size=8] > I/O ports at 8060 [size=4] > I/O ports at 8020 [size=32] > Memory at e9640000 (32-bit, non-prefetchable) [size=2K] > Capabilities: [80] MSI: Enable+ Count=1/1 Maskable- 64bit- > Capabilities: [70] Power Management version 3 > Capabilities: [a8] SATA HBA v1.0 > Capabilities: [b0] PCI Advanced Features > Kernel driver in use: ahci > > Patch not tried yet (do not know / not looked into yet where to change > for this chipset). > > Seems to be a 2.6.39-rcX regression? Well, that very well may be the case. Can you verify that it didn't happen with 2.6.38 and earlier? Rafael
Hello, Can you guys please take a look at bug#34692? It seems to be the same problem. https://bugzilla.kernel.org/show_bug.cgi?id=34692 The offending commit seems to be 270dac35c2. Does reverting it resolves the problem for you guys too? Thanks. -- tejun
On Fri, 13 May 2011 19:39:51 +0200 Tejun Heo <tj@kernel.org> wrote: > Can you guys please take a look at bug#34692? It seems to be the > same problem. > > https://bugzilla.kernel.org/show_bug.cgi?id=34692 > > The offending commit seems to be 270dac35c2. Does reverting it > resolves the problem for you guys too? - Does not happen with 2.6.38.6 - Happens 100% of cases with 2.6.39-rc7 (connect ac adapter or suspend to ram with ac connected and resume with ac disconnected) - Reverting 270dac35c26433d06a89150c51e75ca0181ca7e4 fixes issue
On Friday, May 13, 2011, Michael Leun wrote: > On Fri, 13 May 2011 19:39:51 +0200 > Tejun Heo <tj@kernel.org> wrote: > > > Can you guys please take a look at bug#34692? It seems to be the > > same problem. > > > > https://bugzilla.kernel.org/show_bug.cgi?id=34692 > > > > The offending commit seems to be 270dac35c2. Does reverting it > > resolves the problem for you guys too? > > - Does not happen with 2.6.38.6 > > - Happens 100% of cases with 2.6.39-rc7 (connect ac adapter or suspend > to ram with ac connected and resume with ac disconnected) > > - Reverting 270dac35c26433d06a89150c51e75ca0181ca7e4 fixes issue I can confirm that. Thanks, Rafael
On Friday, May 13, 2011, Tejun Heo wrote: > Hello, > > Can you guys please take a look at bug#34692? It seems to be the same problem. > > https://bugzilla.kernel.org/show_bug.cgi?id=34692 Yes, same issue. > The offending commit seems to be 270dac35c2. Does reverting it > resolves the problem for you guys too? Yes, it does. Thanks, Rafael
On Friday, May 13, 2011, Rafael J. Wysocki wrote: > On Friday, May 13, 2011, Tejun Heo wrote: > > Hello, > > > > Can you guys please take a look at bug#34692? It seems to be the same problem. > > > > https://bugzilla.kernel.org/show_bug.cgi?id=34692 > > Yes, same issue. > > > The offending commit seems to be 270dac35c2. Does reverting it > > resolves the problem for you guys too? > > Yes, it does. Would you object if I asked Linus to revert that commit? Rafael
On 05/13/2011 02:44 PM, Rafael J. Wysocki wrote: > On Friday, May 13, 2011, Rafael J. Wysocki wrote: >> On Friday, May 13, 2011, Tejun Heo wrote: >>> Hello, >>> >>> Can you guys please take a look at bug#34692? It seems to be the same problem. >>> >>> https://bugzilla.kernel.org/show_bug.cgi?id=34692 >> >> Yes, same issue. >> >>> The offending commit seems to be 270dac35c2. Does reverting it >>> resolves the problem for you guys too? >> >> Yes, it does. > > Would you object if I asked Linus to revert that commit? Think that's likely the best solution for now. I'm guessing there's some case where link power management trips up the check that was added. Most or all existing released controllers don't need it so I think Jian could likely try again after 2.6.39.
Hello, On Fri, May 13, 2011 at 10:44:29PM +0200, Rafael J. Wysocki wrote: > On Friday, May 13, 2011, Rafael J. Wysocki wrote: > > On Friday, May 13, 2011, Tejun Heo wrote: > > > Hello, > > > > > > Can you guys please take a look at bug#34692? It seems to be the same problem. > > > > > > https://bugzilla.kernel.org/show_bug.cgi?id=34692 > > > > Yes, same issue. > > > > > The offending commit seems to be 270dac35c2. Does reverting it > > > resolves the problem for you guys too? > > > > Yes, it does. > > Would you object if I asked Linus to revert that commit? Not at all. I think that's the best option at this point. I'll send revert patch right now. Thanks.
Index: linux-2.6/drivers/ata/ahci.c =================================================================== --- linux-2.6.orig/drivers/ata/ahci.c +++ linux-2.6/drivers/ata/ahci.c @@ -193,7 +193,7 @@ static const struct ata_port_info ahci_p [board_ahci_sb700] = /* for SB700 and SB800 */ { AHCI_HFLAGS (AHCI_HFLAG_IGN_SERR_INTERNAL), - .flags = AHCI_FLAG_COMMON, + .flags = AHCI_FLAG_COMMON | ATA_FLAG_NO_DIPM, .pio_mask = ATA_PIO4, .udma_mask = ATA_UDMA6, .port_ops = &ahci_sb600_ops,