lkml.org 
[lkml]   [2005]   [Mar]   [3]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
    /
    Date
    From
    SubjectRe: [PATCH ide-dev-2.6] ide: ide_dma_intr oops fix
    On Thu, Mar 03 2005, Bartlomiej Zolnierkiewicz wrote:
    > On Thu, 03 Mar 2005 15:57:18 +0900, Tejun Heo <htejun@gmail.com> wrote:
    > > Hello, Jens.
    > >
    > > Jens Axboe wrote:
    > > > On Thu, Mar 03 2005, Tejun Heo wrote:
    > > >
    > > >> Hello, Bartlomiej.
    > > >>
    > > >> This patch fixes ide_dma_intr() oops which occurs for TASKFILE ioctl
    > > >>using DMA dataphses. This is against the latest ide-dev-2.6 tree +
    > > >>all your recent 9 patches.
    > > >>
    > > >> Signed-off-by: Tejun Heo <htejun@gmail.com>
    > > >>
    > > >>Index: linux-taskfile-ng/drivers/ide/ide-dma.c
    > > >>===================================================================
    > > >>--- linux-taskfile-ng.orig/drivers/ide/ide-dma.c 2005-03-03 11:59:16.485582413 +0900
    > > >>+++ linux-taskfile-ng/drivers/ide/ide-dma.c 2005-03-03 12:00:07.753376048 +0900
    > > >>@@ -175,10 +175,14 @@ ide_startstop_t ide_dma_intr (ide_drive_
    > > >> if (OK_STAT(stat,DRIVE_READY,drive->bad_wstat|DRQ_STAT)) {
    > > >> if (!dma_stat) {
    > > >> struct request *rq = HWGROUP(drive)->rq;
    > > >>- ide_driver_t *drv;
    > > >>
    > > >>- drv = *(ide_driver_t **)rq->rq_disk->private_data;;
    > > >>- drv->end_request(drive, 1, rq->nr_sectors);
    > > >>+ if (rq->rq_disk) {
    > > >>+ ide_driver_t *drv;
    > > >>+
    > > >>+ drv = *(ide_driver_t **)rq->rq_disk->private_data;;
    > > >>+ drv->end_request(drive, 1, rq->nr_sectors);
    > > >>+ } else
    > > >>+ ide_end_request(drive, 1, rq->nr_sectors);
    > > >> return ide_stopped;
    > > >> }
    > > >> printk(KERN_ERR "%s: dma_intr: bad DMA status (dma_stat=%x)\n",
    > > >
    > > > Why not just set rq_disk for taskfile requests as well, seems a lot
    > > > cleaner than special casing the end_request handling.
    > >
    > > Just because other places were fixed this way and the whole drive
    > > command issue/completion codes are just about to be restructured. Above
    > > code will go away soon. Please consider it a quick fix.
    > >
    > > Thanks.
    >
    > Because struct gendisk is now allocated by device drivers (like in SCSI
    > subsystem) rq_disk can't be set for REQ_DRIVE_TASKFILE requests
    > (for some requests it can be set but better to keep it consistent).

    Seems cleaner to store the driver in the drive structure then, no
    special casing needed.

    --
    Jens Axboe

    -
    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: 2005-03-22 14:10    [W:0.044 / U:62.760 seconds]
    ©2003-2016 Jasper Spaans. hosted at Digital OceanAdvertise on this site