From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:51915) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1bHxh4-0006RS-Ra for qemu-devel@nongnu.org; Tue, 28 Jun 2016 14:25:07 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1bHxh0-0002GY-Eg for qemu-devel@nongnu.org; Tue, 28 Jun 2016 14:25:05 -0400 Received: from 11.mo3.mail-out.ovh.net ([87.98.184.158]:45667) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1bHxh0-0002GI-34 for qemu-devel@nongnu.org; Tue, 28 Jun 2016 14:25:02 -0400 Received: from player772.ha.ovh.net (b7.ovh.net [213.186.33.57]) by mo3.mail-out.ovh.net (Postfix) with ESMTP id 3A89CFF8D59 for ; Tue, 28 Jun 2016 20:25:01 +0200 (CEST) From: =?UTF-8?q?C=C3=A9dric=20Le=20Goater?= Date: Tue, 28 Jun 2016 20:24:23 +0200 Message-Id: <1467138270-32481-3-git-send-email-clg@kaod.org> In-Reply-To: <1467138270-32481-1-git-send-email-clg@kaod.org> References: <1467138270-32481-1-git-send-email-clg@kaod.org> MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: quoted-printable Subject: [Qemu-devel] [PATCH v5 2/9] m25p80: do not put iovec on the stack List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Peter Maydell , Peter Crosthwaite Cc: qemu-devel@nongnu.org, qemu-arm@nongnu.org, kwolf@redhat.com, armbru@redhat.com, Andrew Jeffery , Paolo Bonzini , =?UTF-8?q?C=C3=A9dric=20Le=20Goater?= From: Paolo Bonzini When doing a read-modify-write cycle, QEMU uses the iovec after returning from blk_aio_pwritev. m25p80 puts the iovec on the stack of blk_aio_pwri= tev's caller, which causes trouble in this case. This has been a problem since commit 243e6f6 ("m25p80: Switch to byte-based block access", 2016-05-12) started doing writes at a smaller granularity than 512 bytes. In principle however it could have broken before when using -drive if=3Dmtd,cache=3Dnone on a disk with 4K native sectors. Signed-off-by: Paolo Bonzini Reviewed-by: C=C3=A9dric Le Goater Reviewed-by: Eric Blake Signed-off-by: C=C3=A9dric Le Goater --- hw/block/m25p80.c | 23 ++++++++++++++--------- 1 file changed, 14 insertions(+), 9 deletions(-) diff --git a/hw/block/m25p80.c b/hw/block/m25p80.c index 3cdcfce07ba9..8e4c11573521 100644 --- a/hw/block/m25p80.c +++ b/hw/block/m25p80.c @@ -447,6 +447,11 @@ static inline Manufacturer get_man(Flash *s) =20 static void blk_sync_complete(void *opaque, int ret) { + QEMUIOVector *iov =3D opaque; + + qemu_iovec_destroy(iov); + g_free(iov); + /* do nothing. Masters do not directly interact with the backing sto= re, * only the working copy so no mutexing required. */ @@ -454,31 +459,31 @@ static void blk_sync_complete(void *opaque, int ret= ) =20 static void flash_sync_page(Flash *s, int page) { - QEMUIOVector iov; + QEMUIOVector *iov =3D g_new(QEMUIOVector, 1); =20 if (!s->blk || blk_is_read_only(s->blk)) { return; } =20 - qemu_iovec_init(&iov, 1); - qemu_iovec_add(&iov, s->storage + page * s->pi->page_size, + qemu_iovec_init(iov, 1); + qemu_iovec_add(iov, s->storage + page * s->pi->page_size, s->pi->page_size); - blk_aio_pwritev(s->blk, page * s->pi->page_size, &iov, 0, - blk_sync_complete, NULL); + blk_aio_pwritev(s->blk, page * s->pi->page_size, iov, 0, + blk_sync_complete, iov); } =20 static inline void flash_sync_area(Flash *s, int64_t off, int64_t len) { - QEMUIOVector iov; + QEMUIOVector *iov =3D g_new(QEMUIOVector, 1); =20 if (!s->blk || blk_is_read_only(s->blk)) { return; } =20 assert(!(len % BDRV_SECTOR_SIZE)); - qemu_iovec_init(&iov, 1); - qemu_iovec_add(&iov, s->storage + off, len); - blk_aio_pwritev(s->blk, off, &iov, 0, blk_sync_complete, NULL); + qemu_iovec_init(iov, 1); + qemu_iovec_add(iov, s->storage + off, len); + blk_aio_pwritev(s->blk, off, iov, 0, blk_sync_complete, iov); } =20 static void flash_erase(Flash *s, int offset, FlashCMD cmd) --=20 2.1.4