Re: [2.6.8-rc1-mm1][CD-Burning]kernel BUG at mm/page_alloc.c:796

From: Jens Axboe
Date: Thu Jul 15 2004 - 04:14:29 EST


On Wed, Jul 14 2004, Alexander Gran wrote:
> Hi,
>
> I cannot burn cd's with 2.6.8-rc1-mm1.
> System is Fujustu Siemens Celsius, detailed info below.
> I hope it is not NVidias fault, but I don't think so.
> I gonna check older kernels now.
>
> root@ipt-cax-ws1: /root >cdrecord dev=/dev/hdc --checkdrive
> results in:
> ------------[ cut here ]------------
> kernel BUG at mm/page_alloc.c:796!
> invalid operand: 0000 [#1]
> PREEMPT SMP
> Modules linked in: nvidia thermal processor fan button battery ac
> CPU: 0
> EIP: 0060:[<c01421ef>] Tainted: P VLI
> EFLAGS: 00010246 (2.6.8-rc1-mm1)
> EIP is at __free_pages+0x3e/0x48
> eax: ffffffff ebx: c2197400 ecx: c16c94a0 edx: 00000000
> esi: c2191600 edi: 00000001 ebp: c2191680 esp: f52ddd34
> ds: 007b es: 007b ss: 0068
> Process cdrecord (pid: 1968, threadinfo=f52dc000 task=f5f562b0)
> Stack: c2197400 c2191600 c0164b78 c2191600 f64a5000 00000100 fffffff2 00000000
> 00079963 c2191680 00000000 c0392dcb c2191680 f5223eac c0396b2c f5223eac
> c2191680 00000100 00000100 21000046 01000000 00000000 f64d7930 00000000
> Call Trace:
> [<c0164b78>] bio_uncopy_user+0x84/0xa9
> [<c0392dcb>] blk_rq_unmap_user+0x37/0x4c
> [<c0396b2c>] sg_io+0x296/0x310
> [<c03971dc>] scsi_cmd_ioctl+0x363/0x487

Can you check if this patch works for you? Andrew, this also fixes a
problem in the bouncing for io errors (it needs to free the pages and
clear the BIO_UPTODATE flag, not set it. it's already set. passing -EIO
to bio_endio() takes care of that).

diff -urp -X /home/axboe/cdrom/exclude /opt/kernel/linux-2.6.8-rc1-mm1/drivers/block/ll_rw_blk.c linux-2.6.8-rc1-mm1/drivers/block/ll_rw_blk.c
--- /opt/kernel/linux-2.6.8-rc1-mm1/drivers/block/ll_rw_blk.c 2004-07-15 10:41:31.678544355 +0200
+++ linux-2.6.8-rc1-mm1/drivers/block/ll_rw_blk.c 2004-07-15 10:43:09.096987402 +0200
@@ -1853,6 +1853,12 @@ EXPORT_SYMBOL(blk_insert_request);
*
* A matching blk_rq_unmap_user() must be issued at the end of io, while
* still in process context.
+ *
+ * Note: The mapped bio may need to be bounced through blk_queue_bounce()
+ * before being submitted to the device, as pages mapped may be out of
+ * reach. It's the callers responsibility to make sure this happens. The
+ * original bio must be passed back in to blk_rq_unmap_user() for proper
+ * unmapping.
*/
struct request *blk_rq_map_user(request_queue_t *q, int rw, void __user *ubuf,
unsigned int len)
diff -urp -X /home/axboe/cdrom/exclude /opt/kernel/linux-2.6.8-rc1-mm1/drivers/block/scsi_ioctl.c linux-2.6.8-rc1-mm1/drivers/block/scsi_ioctl.c
--- /opt/kernel/linux-2.6.8-rc1-mm1/drivers/block/scsi_ioctl.c 2004-07-15 10:41:31.762535252 +0200
+++ linux-2.6.8-rc1-mm1/drivers/block/scsi_ioctl.c 2004-07-15 10:43:09.092987835 +0200
@@ -170,6 +170,13 @@ static int sg_io(request_queue_t *q, str
rq->flags |= REQ_BLOCK_PC;
bio = rq->bio;

+ /*
+ * bounce this after holding a reference to the original bio, it's
+ * needed for proper unmapping
+ */
+ if (rq->bio)
+ blk_queue_bounce(q, &rq->bio);
+
rq->timeout = (hdr->timeout * HZ) / 1000;
if (!rq->timeout)
rq->timeout = q->sg_timeout;
diff -urp -X /home/axboe/cdrom/exclude /opt/kernel/linux-2.6.8-rc1-mm1/drivers/cdrom/cdrom.c linux-2.6.8-rc1-mm1/drivers/cdrom/cdrom.c
--- /opt/kernel/linux-2.6.8-rc1-mm1/drivers/cdrom/cdrom.c 2004-07-15 10:41:32.994401746 +0200
+++ linux-2.6.8-rc1-mm1/drivers/cdrom/cdrom.c 2004-07-15 10:43:09.094987619 +0200
@@ -2082,6 +2082,9 @@ static int cdrom_read_cdda_bpc(struct cd
rq->timeout = 60 * HZ;
bio = rq->bio;

+ if (rq->bio)
+ blk_queue_bounce(q, &rq->bio);
+
if (blk_execute_rq(q, cdi->disk, rq)) {
struct request_sense *s = rq->sense;
ret = -EIO;
diff -urp -X /home/axboe/cdrom/exclude /opt/kernel/linux-2.6.8-rc1-mm1/fs/bio.c linux-2.6.8-rc1-mm1/fs/bio.c
--- /opt/kernel/linux-2.6.8-rc1-mm1/fs/bio.c 2004-07-15 10:41:44.105197605 +0200
+++ linux-2.6.8-rc1-mm1/fs/bio.c 2004-07-15 10:43:09.091987944 +0200
@@ -557,7 +557,6 @@ static struct bio *__bio_map_user(reques
bio->bi_rw |= (1 << BIO_RW);

bio->bi_flags |= (1 << BIO_USER_MAPPED);
- blk_queue_bounce(q, &bio);
return bio;
out:
kfree(pages);
diff -urp -X /home/axboe/cdrom/exclude /opt/kernel/linux-2.6.8-rc1-mm1/mm/highmem.c linux-2.6.8-rc1-mm1/mm/highmem.c
--- /opt/kernel/linux-2.6.8-rc1-mm1/mm/highmem.c 2004-07-15 10:41:50.046553716 +0200
+++ linux-2.6.8-rc1-mm1/mm/highmem.c 2004-07-15 10:43:09.090988052 +0200
@@ -308,12 +308,10 @@ static void bounce_end_io(struct bio *bi
{
struct bio *bio_orig = bio->bi_private;
struct bio_vec *bvec, *org_vec;
- int i;
+ int i, err = 0;

if (!test_bit(BIO_UPTODATE, &bio->bi_flags))
- goto out_eio;
-
- set_bit(BIO_UPTODATE, &bio_orig->bi_flags);
+ err = -EIO;

/*
* free up bounce indirect pages used
@@ -326,8 +324,7 @@ static void bounce_end_io(struct bio *bi
mempool_free(bvec->bv_page, pool);
}

-out_eio:
- bio_endio(bio_orig, bio_orig->bi_size, 0);
+ bio_endio(bio_orig, bio_orig->bi_size, err);
bio_put(bio);
}


--
Jens Axboe

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