From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:38635) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1clVbw-0005Zp-Ez for qemu-devel@nongnu.org; Wed, 08 Mar 2017 02:02:13 -0500 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1clVbv-0000v4-LE for qemu-devel@nongnu.org; Wed, 08 Mar 2017 02:02:12 -0500 From: Fam Zheng Date: Wed, 8 Mar 2017 15:01:59 +0800 Message-Id: <20170308070159.19487-1-famz@redhat.com> Subject: [Qemu-devel] [PATCH v2] file-posix: Incoporate max_segments in block limit List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: qemu-devel@nongnu.org Cc: qemu-block@nongnu.org, Kevin Wolf , Max Reitz , pbonzini@redhat.com, eblake@redhat.com Linux exposes a separate limit, /sys/block/.../queue/max_segments, which in the worst case can be more restrictive than BLKSECTGET (as they are two different things). Similar to the BLKSECTGET story, guests don't see this limit and send big requests will get -EINVAL error on SG_IO. Lean on the safer side to clamp max_transfer according to max_segments and page size, because in the end what host HBA gets is the mapped host pages rather than a guest buffer. Signed-off-by: Fam Zheng --- v2: Use /sys/dev/block/MAJOR:MINOR/queue/max_segments. [Paolo] --- block/file-posix.c | 47 +++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 47 insertions(+) diff --git a/block/file-posix.c b/block/file-posix.c index 4de1abd..c4c0663 100644 --- a/block/file-posix.c +++ b/block/file-posix.c @@ -668,6 +668,48 @@ static int hdev_get_max_transfer_length(BlockDriverState *bs, int fd) #endif } +static int hdev_get_max_segments(const struct stat *st) +{ +#ifdef CONFIG_LINUX + char buf[32]; + const char *end; + char *sysfspath; + int ret; + int fd = -1; + long max_segments; + + sysfspath = g_strdup_printf("/sys/dev/block/%u:%u/queue/max_segments", + major(st->st_rdev), minor(st->st_rdev)); + fd = open(sysfspath, O_RDONLY); + if (fd == -1) { + ret = -errno; + goto out; + } + do { + ret = read(fd, buf, sizeof(buf)); + } while (ret == -1 && errno == EINTR); + if (ret < 0) { + ret = -errno; + goto out; + } else if (ret == 0) { + ret = -EIO; + goto out; + } + buf[ret] = 0; + /* The file is ended with '\n', pass 'end' to accept that. */ + ret = qemu_strtol(buf, &end, 10, &max_segments); + if (ret == 0 && end && *end == '\n') { + ret = max_segments; + } + +out: + g_free(sysfspath); + return ret; +#else + return -ENOTSUP; +#endif +} + static void raw_refresh_limits(BlockDriverState *bs, Error **errp) { BDRVRawState *s = bs->opaque; @@ -679,6 +721,11 @@ static void raw_refresh_limits(BlockDriverState *bs, Error **errp) if (ret > 0 && ret <= BDRV_REQUEST_MAX_BYTES) { bs->bl.max_transfer = pow2floor(ret); } + ret = hdev_get_max_segments(&st); + if (ret > 0) { + bs->bl.max_transfer = MIN(bs->bl.max_transfer, + ret * getpagesize()); + } } } -- 2.9.3