The patch titled
Subject: drivers/rapidio/rio_cm.c: avoid GFP_KERNEL in atomic context
has been added to the -mm tree. Its filename is
drivers-rapidio-rio_cmc-avoid-gfp_kernel-in-atomic-context.patch
This patch should soon appear at
http://ozlabs.org/~akpm/mmots/broken-out/drivers-rapidio-rio_cmc-avoid-gfp_kernel-in-atomic-context.patch
and later at
http://ozlabs.org/~akpm/mmotm/broken-out/drivers-rapidio-rio_cmc-avoid-gfp_kernel-in-atomic-context.patch
Before you just go and hit "reply", please:
a) Consider who else should be cc'ed
b) Prefer to cc a suitable mailing list as well
c) Ideally: find the original patch on the mailing list and do a
reply-to-all to that, adding suitable additional cc's
*** Remember to use Documentation/SubmitChecklist when testing your code ***
The -mm tree is included into linux-next and is updated
there every 3-4 working days
------------------------------------------------------
From: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
Subject: drivers/rapidio/rio_cm.c: avoid GFP_KERNEL in atomic context
riocm_send_close() is called from rio_cm_shutdown() under
spin_lock_bh(idr_lock), but riocm_send_close() uses a GFP_KERNEL
allocation.
Fix this by making local `hdr' a local variable rather than obtaining it
via kmalloc().
Found by Linux Driver Verification project (linuxtesting.org).
Link: http://lkml.kernel.org/r/1473453815-15914-1-git-send-email-khoroshilov@xxxxxxxxx
Reported-by: Alexey Khoroshilov <khoroshilov@xxxxxxxxx>
Cc: Alexandre Bounine <alexandre.bounine@xxxxxxx>
Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
---
drivers/rapidio/rio_cm.c | 28 ++++++++++++----------------
1 file changed, 12 insertions(+), 16 deletions(-)
diff -puN drivers/rapidio/rio_cm.c~drivers-rapidio-rio_cmc-avoid-gfp_kernel-in-atomic-context drivers/rapidio/rio_cm.c
--- a/drivers/rapidio/rio_cm.c~drivers-rapidio-rio_cmc-avoid-gfp_kernel-in-atomic-context
+++ a/drivers/rapidio/rio_cm.c
@@ -1397,36 +1397,32 @@ static void riocm_ch_free(struct kref *r
static int riocm_send_close(struct rio_channel *ch)
{
- struct rio_ch_chan_hdr *hdr;
+ struct rio_ch_chan_hdr hdr;
int ret;
/*
* Send CH_CLOSE notification to the remote RapidIO device
*/
- hdr = kzalloc(sizeof(*hdr), GFP_KERNEL);
- if (hdr == NULL)
- return -ENOMEM;
-
- hdr->bhdr.src_id = htonl(ch->loc_destid);
- hdr->bhdr.dst_id = htonl(ch->rem_destid);
- hdr->bhdr.src_mbox = cmbox;
- hdr->bhdr.dst_mbox = cmbox;
- hdr->bhdr.type = RIO_CM_CHAN;
- hdr->ch_op = CM_CONN_CLOSE;
- hdr->dst_ch = htons(ch->rem_channel);
- hdr->src_ch = htons(ch->id);
+ memset(&hdr, 0, sizeof(hdr));
+ hdr.bhdr.src_id = htonl(ch->loc_destid);
+ hdr.bhdr.dst_id = htonl(ch->rem_destid);
+ hdr.bhdr.src_mbox = cmbox;
+ hdr.bhdr.dst_mbox = cmbox;
+ hdr.bhdr.type = RIO_CM_CHAN;
+ hdr.ch_op = CM_CONN_CLOSE;
+ hdr.dst_ch = htons(ch->rem_channel);
+ hdr.src_ch = htons(ch->id);
/* ATTN: the function call below relies on the fact that underlying
* add_outb_message() routine copies TX data into its internal transfer
* buffer. Needs to be reviewed if switched to direct buffer mode.
*/
- ret = riocm_post_send(ch->cmdev, ch->rdev, hdr, sizeof(*hdr));
+ ret = riocm_post_send(ch->cmdev, ch->rdev, &hdr, sizeof(hdr));
if (ret == -EBUSY && !riocm_queue_req(ch->cmdev, ch->rdev,
- hdr, sizeof(*hdr)))
+ &hdr, sizeof(hdr)))
return 0;
- kfree(hdr);
if (ret)
riocm_error("ch(%d) send CLOSE failed (ret=%d)", ch->id, ret);
_
Patches currently in -mm which might be from akpm@xxxxxxxxxxxxxxxxxxxx are
arch-alpha-kernel-systblss-remove-debug-check.patch
i-need-old-gcc.patch
drivers-rapidio-rio_cmc-avoid-gfp_kernel-in-atomic-context.patch
arm-arch-arm-include-asm-pageh-needs-personalityh.patch
mm.patch
mm-vmalloc-fix-align-value-calculation-error-fix.patch
mm-vmalloc-fix-align-value-calculation-error-v2-fix.patch
mm-vmalloc-fix-align-value-calculation-error-v2-fix-fix.patch
mm-compaction-make-whole_zone-flag-ignore-cached-scanner-positions-checkpatch-fixes.patch
mm-swap-add-swap_cluster_list-checkpatch-fixes.patch
mm-check-that-we-havent-used-more-than-32-bits-in-address_spaceflags.patch
mm-mlock-check-against-vma-for-actual-mlock-size-fix.patch
seq-proc-modify-seq_put_decimal_ll-to-take-a-const-char-not-char-fix.patch
relay-use-irq_work-instead-of-plain-timer-for-deferred-wakeup-checkpatch-fixes.patch
ipc-msg-avoid-waking-sender-upon-full-queue-checkpatch-fixes.patch
linux-next-rejects.patch
drivers-net-wireless-intel-iwlwifi-dvm-calibc-fix-min-warning.patch
include-linux-mlx5-deviceh-kill-build_bug_ons.patch
hung_task-allow-hung_task_panic-when-hung_task_warnings-is-0-fix.patch
kernel-forkc-export-kernel_thread-to-modules.patch
slab-leaks3-default-y.patch
--
To unsubscribe from this list: send the line "unsubscribe mm-commits" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
