via82cxxx resume failure.
On resume from sleep, via_set_speed() doesn't reinstate the correct DMA mode, because it thinks the drive is already configured correctly. This one-line hack is sufficient to make it refrain from dying a horrible death immediately after resume, but presumably has other problems...
===== drivers/ide/pci/via82cxxx.c 1.24 vs edited =====
drivers/ide/pci/via82cxxx.c +1 -1
@@ -328,7 +328,7 @@struct ide_timing t, p;unsigned int T ,UT ; -if (speed !=XFER_PIO_SLOW && speed != drive->current_speed) +// if (speed != XFER_PIO_SLOW && speed != drive->current_speed) if (ide_config_drive_speed (drive, speed)) printk(KERN_WARNING "ide%d: Drive %d didn't " "accept speed setting. Oh, well.\n",
Signature · 1 line
dwmw2
Sorry for the delay.
Quoted 4 lines · David Woodhouse
On resume from sleep,
via_set_speed()doesn't reinstate the correct DMA mode, because it thinks the drive is already configured correctly. This one-line hack is sufficient to make it refrain from dying a horrible death immediately after resume, but presumably has other problems...
I applied this to libata-dev so it gets some testing in -mm.
Quoted 13 lines
=====
drivers/ide/pci/via82cxxx.c1.24 vs edited =====drivers/ide/pci/via82cxxx.c Tue Oct 26 22:48:59 2004 renamed from drivers/ide/pci/via82cxxx.c Mon Aug 9 18:00:46 2004 +1
-1@@ -328,7 +328,7 @@ struct ide_timing t, p; unsigned int T, UT; - if (speed != XFER_PIO_SLOW && speed != drive->current_speed) + // if (speed != XFER_PIO_SLOW && speed != drive->current_speed) if (ide_config_drive_speed(drive, speed)) printk(KERN_WARNING "ide%d: Drive %d didn't " "accept speed setting. Oh, well.\n",
Quoted 11 lines · Bartlomiej Zolnierkiewicz
Sorry for the delay.
Quoted 4 lines · David Woodhouse
On resume from sleep,
via_set_speed()doesn't reinstate the correct DMA mode, because it thinks the drive is already configured correctly. This one-line hack is sufficient to make it refrain from dying a horrible death immediately after resume, but presumably has other problems...I applied this to libata-dev so it gets some testing in -mm.
Chuckle -- you mean ide-dev, presumably :)
Jeff
Quoted 11 lines · Jeff Garzik
Quoted 10 lines · Bartlomiej Zolnierkiewicz
Sorry for the delay.
Quoted 4 lines · David Woodhouse
On resume from sleep,
via_set_speed()doesn't reinstate the correct DMA mode, because it thinks the drive is already configured correctly. This one-line hack is sufficient to make it refrain from dying a horrible death immediately after resume, but presumably has other problems...
Quoted 3 lines
Quoted 1 line
I applied this to libata-dev so it gets some testing in -mm.
Chuckle -- you mean ide-dev, presumably :)
yes, obviously :)