Messages in this thread Patch in this message | | | Date | Tue, 31 Mar 2009 09:45:08 +0200 | From | Borislav Petkov <> | Subject | Re: Bug in 2.6.29 ide-cd: Kernel freeze: bisected + unacceptable workaround |
| |
Hi,
> > @Bart: can you take at look at this. Somehow, if the device is behind a > > PCI IDE controller, the order of issuing the command and enabling DMA is > > important. Seen something like that before? > > Spec is unclear on the ordering but empirically it seems that some hosts > may need packet command to start DMA. I think that we should proceed with > your patch (please repost with patch description) and also apply the same > change to the rest of ATAPI devices in a subsequent patch.
Why two patches, do we really need to differentiate between ide-cd and other ATAPI devices? Besides the patch is so simple: 4 lines moved up. Anyway, I lightly tested it with ide-cd and ide-floppy and both seem to take it ok. Also, if I remember correctly, the original ide-cd behavior was to issue the command and _then_ start DMA so we're back to that. I guess now we should be concerned whether the other ATAPI devices can handle the reverse situation where you first issue a command and _then_ start DMA.. Hmm...
-- From: Borislav Petkov <petkovbb@gmail.com> Date: Tue, 31 Mar 2009 09:36:44 +0200 Subject: [PATCH] ide-atapi: start DMA after issuing a packet command MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: 8bit
Apparently¹, some ATAPI devices want to see the packet command first before enabling DMA otherwise they simply hang indefinitely. Reorder the two steps and start DMA only after having issued the command first.
[1] http://marc.info/?l=linux-kernel&m=123835520317235&w=2
Signed-off-by: Borislav Petkov <petkovbb@gmail.com> --- drivers/ide/ide-atapi.c | 8 ++++---- 1 files changed, 4 insertions(+), 4 deletions(-)
diff --git a/drivers/ide/ide-atapi.c b/drivers/ide/ide-atapi.c index c139cc7..3e43b88 100644 --- a/drivers/ide/ide-atapi.c +++ b/drivers/ide/ide-atapi.c @@ -563,6 +563,10 @@ static ide_startstop_t ide_transfer_pc(ide_drive_t *drive) : ide_pc_intr), timeout); + /* Send the actual packet */ + if ((drive->atapi_flags & IDE_AFLAG_ZIP_DRIVE) == 0) + hwif->tp_ops->output_data(drive, NULL, rq->cmd, cmd_len); + /* Begin DMA, if necessary */ if (dev_is_idecd(drive)) { if (drive->dma) @@ -574,10 +578,6 @@ static ide_startstop_t ide_transfer_pc(ide_drive_t *drive) } } - /* Send the actual packet */ - if ((drive->atapi_flags & IDE_AFLAG_ZIP_DRIVE) == 0) - hwif->tp_ops->output_data(drive, NULL, rq->cmd, cmd_len); - return ide_started; } -- 1.6.2.1
-- Regards/Gruss, Boris. -- 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/
| |