The following changes since commit 23919ddfd56135cad3cb468a8f54d5a595f024f4: Merge remote-tracking branch 'remotes/aperard/tags/pull-xen-20190827' into staging (2019-08-27 15:52:36 +0100) are available in the Git repository at: https://github.com/XanClic/qemu.git tags/pull-block-2019-08-27 for you to fetch changes up to bb043c056cffcc2f3ce88bfdaf2e76e455c09e2c: iotests: Unify cache mode quoting (2019-08-27 19:48:44 +0200) ---------------------------------------------------------------- Block patches: - qemu-io now accepts a file to read a write pattern from - Ensure that raw files have their first block allocated so we can probe the O_DIRECT alignment if necessary - Various fixes ---------------------------------------------------------------- Denis Plotnikov (1): qemu-io: add pattern file for write command Max Reitz (7): iotests: Fix _filter_img_create() vmdk: Use bdrv_dirname() for relative extent paths iotests: Keep testing broken relative extent paths vmdk: Reject invalid compressed writes iotests: Disable broken streamOptimized tests iotests: Disable 110 for vmdk.twoGbMaxExtentSparse iotests: Disable 126 for flat vmdk subformats Nir Soffer (3): block: posix: Always allocate the first block iotests: Test allocate_first_block() with O_DIRECT iotests: Unify cache mode quoting Stefan Hajnoczi (1): file-posix: fix request_alignment typo Thomas Huth (2): iotests: Check for enabled drivers before testing them tests/check-block: Skip iotests when sanitizers are enabled Vladimir Sementsov-Ogievskiy (1): block: fix permission update in bdrv_replace_node block.c | 5 +- block/file-posix.c | 53 +++++++++- block/vmdk.c | 64 ++++++++---- qemu-io-cmds.c | 99 +++++++++++++++++-- tests/check-block.sh | 5 + tests/qemu-iotests/002 | 1 + tests/qemu-iotests/003 | 1 + tests/qemu-iotests/005 | 3 +- tests/qemu-iotests/009 | 1 + tests/qemu-iotests/010 | 1 + tests/qemu-iotests/011 | 1 + tests/qemu-iotests/017 | 3 +- tests/qemu-iotests/018 | 3 +- tests/qemu-iotests/019 | 3 +- tests/qemu-iotests/020 | 3 +- tests/qemu-iotests/026 | 4 +- tests/qemu-iotests/027 | 1 + tests/qemu-iotests/032 | 1 + tests/qemu-iotests/033 | 1 + tests/qemu-iotests/034 | 3 +- tests/qemu-iotests/037 | 3 +- tests/qemu-iotests/039 | 4 +- tests/qemu-iotests/052 | 2 +- tests/qemu-iotests/059 | 34 ++++++- tests/qemu-iotests/059.out | 26 +++-- tests/qemu-iotests/063 | 3 +- tests/qemu-iotests/071 | 1 + tests/qemu-iotests/072 | 1 + tests/qemu-iotests/081 | 4 +- tests/qemu-iotests/091 | 4 +- tests/qemu-iotests/099 | 1 + tests/qemu-iotests/105 | 3 +- tests/qemu-iotests/110 | 3 +- tests/qemu-iotests/120 | 1 + tests/qemu-iotests/126 | 2 + tests/qemu-iotests/{150.out => 150.out.qcow2} | 0 tests/qemu-iotests/150.out.raw | 12 +++ tests/qemu-iotests/162 | 4 +- tests/qemu-iotests/175 | 47 +++++++-- tests/qemu-iotests/175.out | 16 ++- tests/qemu-iotests/178.out.qcow2 | 4 +- tests/qemu-iotests/184 | 1 + tests/qemu-iotests/186 | 1 + tests/qemu-iotests/197 | 1 + tests/qemu-iotests/215 | 1 + tests/qemu-iotests/221.out | 12 ++- tests/qemu-iotests/251 | 1 + tests/qemu-iotests/253.out | 12 ++- tests/qemu-iotests/common.filter | 4 +- tests/qemu-iotests/common.rc | 14 +++ 50 files changed, 391 insertions(+), 87 deletions(-) rename tests/qemu-iotests/{150.out => 150.out.qcow2} (100%) create mode 100644 tests/qemu-iotests/150.out.raw -- 2.21.0 From: Denis Plotnikov <dplotnikov@virtuozzo.com> The patch allows to provide a pattern file for write command. There was no similar ability before. Signed-off-by: Denis Plotnikov <dplotnikov@virtuozzo.com> Message-id: 20190820164616.4072-1-dplotnikov@virtuozzo.com Reviewed-by: Eric Blake <eblake@redhat.com> [mreitz: Keep optstring in alphabetical order] Signed-off-by: Max Reitz <mreitz@redhat.com> --- qemu-io-cmds.c | 99 +++++++++++++++++++++++++++++++++++++++++++++++--- 1 file changed, 93 insertions(+), 6 deletions(-) diff --git a/qemu-io-cmds.c b/qemu-io-cmds.c index XXXXXXX..XXXXXXX 100644 --- a/qemu-io-cmds.c +++ b/qemu-io-cmds.c @@ -XXX,XX +XXX,XX @@ static void qemu_io_free(void *p) qemu_vfree(p); } +/* + * qemu_io_alloc_from_file() + * + * Allocates the buffer and populates it with the content of the given file + * up to @len bytes. If the file length is less than @len, then the buffer + * is populated with the file content cyclically. + * + * @blk - the block backend where the buffer content is going to be written to + * @len - the buffer length + * @file_name - the file to read the content from + * + * Returns: the buffer pointer on success + * NULL on error + */ +static void *qemu_io_alloc_from_file(BlockBackend *blk, size_t len, + const char *file_name) +{ + char *buf, *buf_origin; + FILE *f = fopen(file_name, "r"); + int pattern_len; + + if (!f) { + perror(file_name); + return NULL; + } + + if (qemuio_misalign) { + len += MISALIGN_OFFSET; + } + + buf_origin = buf = blk_blockalign(blk, len); + + if (qemuio_misalign) { + buf_origin += MISALIGN_OFFSET; + buf += MISALIGN_OFFSET; + len -= MISALIGN_OFFSET; + } + + pattern_len = fread(buf_origin, 1, len, f); + + if (ferror(f)) { + perror(file_name); + goto error; + } + + if (pattern_len == 0) { + fprintf(stderr, "%s: file is empty\n", file_name); + goto error; + } + + fclose(f); + + if (len > pattern_len) { + len -= pattern_len; + buf += pattern_len; + + while (len > 0) { + size_t len_to_copy = MIN(pattern_len, len); + + memcpy(buf, buf_origin, len_to_copy); + + len -= len_to_copy; + buf += len_to_copy; + } + } + + return buf_origin; + +error: + qemu_io_free(buf_origin); + return NULL; +} + static void dump_buffer(const void *buffer, int64_t offset, int64_t len) { uint64_t i; @@ -XXX,XX +XXX,XX @@ static void write_help(void) " -n, -- with -z, don't allow slow fallback\n" " -p, -- ignored for backwards compatibility\n" " -P, -- use different pattern to fill file\n" +" -s, -- use a pattern file to fill the write buffer\n" " -C, -- report statistics in a machine parsable format\n" " -q, -- quiet mode, do not show I/O statistics\n" " -u, -- with -z, allow unmapping\n" @@ -XXX,XX +XXX,XX @@ static const cmdinfo_t write_cmd = { .perm = BLK_PERM_WRITE, .argmin = 2, .argmax = -1, - .args = "[-bcCfnquz] [-P pattern] off len", + .args = "[-bcCfnquz] [-P pattern | -s source_file] off len", .oneline = "writes a number of bytes at a specified offset", .help = write_help, }; @@ -XXX,XX +XXX,XX @@ static int write_f(BlockBackend *blk, int argc, char **argv) { struct timespec t1, t2; bool Cflag = false, qflag = false, bflag = false; - bool Pflag = false, zflag = false, cflag = false; + bool Pflag = false, zflag = false, cflag = false, sflag = false; int flags = 0; int c, cnt, ret; char *buf = NULL; @@ -XXX,XX +XXX,XX @@ static int write_f(BlockBackend *blk, int argc, char **argv) /* Some compilers get confused and warn if this is not initialized. */ int64_t total = 0; int pattern = 0xcd; + const char *file_name = NULL; - while ((c = getopt(argc, argv, "bcCfnpP:quz")) != -1) { + while ((c = getopt(argc, argv, "bcCfnpP:qs:uz")) != -1) { switch (c) { case 'b': bflag = true; @@ -XXX,XX +XXX,XX @@ static int write_f(BlockBackend *blk, int argc, char **argv) case 'q': qflag = true; break; + case 's': + sflag = true; + file_name = optarg; + break; case 'u': flags |= BDRV_REQ_MAY_UNMAP; break; @@ -XXX,XX +XXX,XX @@ static int write_f(BlockBackend *blk, int argc, char **argv) return -EINVAL; } - if (zflag && Pflag) { - printf("-z and -P cannot be specified at the same time\n"); + if (zflag + Pflag + sflag > 1) { + printf("Only one of -z, -P, and -s " + "can be specified at the same time\n"); return -EINVAL; } @@ -XXX,XX +XXX,XX @@ static int write_f(BlockBackend *blk, int argc, char **argv) } if (!zflag) { - buf = qemu_io_alloc(blk, count, pattern); + if (sflag) { + buf = qemu_io_alloc_from_file(blk, count, file_name); + if (!buf) { + return -EINVAL; + } + } else { + buf = qemu_io_alloc(blk, count, pattern); + } } clock_gettime(CLOCK_MONOTONIC, &t1); -- 2.21.0 From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> It's wrong to OR shared permissions. It may lead to crash on further permission updates. Also, no needs to consider previously calculated permissions, as at this point we already bind all new parents and bdrv_get_cumulative_perm result is enough. So fix the bug by just set permissions by bdrv_get_cumulative_perm result. Bug was introduced in long ago 234ac1a9025, in 2.9. Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Message-id: 20190824100740.61635-1-vsementsov@virtuozzo.com Signed-off-by: Max Reitz <mreitz@redhat.com> --- block.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) diff --git a/block.c b/block.c index XXXXXXX..XXXXXXX 100644 --- a/block.c +++ b/block.c @@ -XXX,XX +XXX,XX @@ void bdrv_replace_node(BlockDriverState *from, BlockDriverState *to, { BdrvChild *c, *next; GSList *list = NULL, *p; - uint64_t old_perm, old_shared; uint64_t perm = 0, shared = BLK_PERM_ALL; int ret; @@ -XXX,XX +XXX,XX @@ void bdrv_replace_node(BlockDriverState *from, BlockDriverState *to, bdrv_unref(from); } - bdrv_get_cumulative_perm(to, &old_perm, &old_shared); - bdrv_set_perm(to, old_perm | perm, old_shared | shared); + bdrv_get_cumulative_perm(to, &perm, &shared); + bdrv_set_perm(to, perm, shared); out: g_slist_free(list); -- 2.21.0 From: Nir Soffer <nirsof@gmail.com> When creating an image with preallocation "off" or "falloc", the first block of the image is typically not allocated. When using Gluster storage backed by XFS filesystem, reading this block using direct I/O succeeds regardless of request length, fooling alignment detection. In this case we fallback to a safe value (4096) instead of the optimal value (512), which may lead to unneeded data copying when aligning requests. Allocating the first block avoids the fallback. Since we allocate the first block even with preallocation=off, we no longer create images with zero disk size: $ ./qemu-img create -f raw test.raw 1g Formatting 'test.raw', fmt=raw size=1073741824 $ ls -lhs test.raw 4.0K -rw-r--r--. 1 nsoffer nsoffer 1.0G Aug 16 23:48 test.raw And converting the image requires additional cluster: $ ./qemu-img measure -f raw -O qcow2 test.raw required size: 458752 fully allocated size: 1074135040 When using format like vmdk with multiple files per image, we allocate one block per file: $ ./qemu-img create -f vmdk -o subformat=twoGbMaxExtentFlat test.vmdk 4g Formatting 'test.vmdk', fmt=vmdk size=4294967296 compat6=off hwversion=undefined subformat=twoGbMaxExtentFlat $ ls -lhs test*.vmdk 4.0K -rw-r--r--. 1 nsoffer nsoffer 2.0G Aug 27 03:23 test-f001.vmdk 4.0K -rw-r--r--. 1 nsoffer nsoffer 2.0G Aug 27 03:23 test-f002.vmdk 4.0K -rw-r--r--. 1 nsoffer nsoffer 353 Aug 27 03:23 test.vmdk I did quick performance test for copying disks with qemu-img convert to new raw target image to Gluster storage with sector size of 512 bytes: for i in $(seq 10); do rm -f dst.raw sleep 10 time ./qemu-img convert -f raw -O raw -t none -T none src.raw dst.raw done Here is a table comparing the total time spent: Type Before(s) After(s) Diff(%) --------------------------------------- real 530.028 469.123 -11.4 user 17.204 10.768 -37.4 sys 17.881 7.011 -60.7 We can see very clear improvement in CPU usage. Signed-off-by: Nir Soffer <nsoffer@redhat.com> Message-id: 20190827010528.8818-2-nsoffer@redhat.com Reviewed-by: Max Reitz <mreitz@redhat.com> Signed-off-by: Max Reitz <mreitz@redhat.com> --- block/file-posix.c | 51 +++++++++++++++++++ tests/qemu-iotests/059.out | 2 +- tests/qemu-iotests/{150.out => 150.out.qcow2} | 0 tests/qemu-iotests/150.out.raw | 12 +++++ tests/qemu-iotests/175 | 19 ++++--- tests/qemu-iotests/175.out | 8 +-- tests/qemu-iotests/178.out.qcow2 | 4 +- tests/qemu-iotests/221.out | 12 +++-- tests/qemu-iotests/253.out | 12 +++-- 9 files changed, 99 insertions(+), 21 deletions(-) rename tests/qemu-iotests/{150.out => 150.out.qcow2} (100%) create mode 100644 tests/qemu-iotests/150.out.raw diff --git a/block/file-posix.c b/block/file-posix.c index XXXXXXX..XXXXXXX 100644 --- a/block/file-posix.c +++ b/block/file-posix.c @@ -XXX,XX +XXX,XX @@ static int handle_aiocb_discard(void *opaque) return ret; } +/* + * Help alignment probing by allocating the first block. + * + * When reading with direct I/O from unallocated area on Gluster backed by XFS, + * reading succeeds regardless of request length. In this case we fallback to + * safe alignment which is not optimal. Allocating the first block avoids this + * fallback. + * + * fd may be opened with O_DIRECT, but we don't know the buffer alignment or + * request alignment, so we use safe values. + * + * Returns: 0 on success, -errno on failure. Since this is an optimization, + * caller may ignore failures. + */ +static int allocate_first_block(int fd, size_t max_size) +{ + size_t write_size = (max_size < MAX_BLOCKSIZE) + ? BDRV_SECTOR_SIZE + : MAX_BLOCKSIZE; + size_t max_align = MAX(MAX_BLOCKSIZE, getpagesize()); + void *buf; + ssize_t n; + int ret; + + buf = qemu_memalign(max_align, write_size); + memset(buf, 0, write_size); + + do { + n = pwrite(fd, buf, write_size, 0); + } while (n == -1 && errno == EINTR); + + ret = (n == -1) ? -errno : 0; + + qemu_vfree(buf); + return ret; +} + static int handle_aiocb_truncate(void *opaque) { RawPosixAIOData *aiocb = opaque; @@ -XXX,XX +XXX,XX @@ static int handle_aiocb_truncate(void *opaque) /* posix_fallocate() doesn't set errno. */ error_setg_errno(errp, -result, "Could not preallocate new data"); + } else if (current_length == 0) { + /* + * posix_fallocate() uses fallocate() if the filesystem + * supports it, or fallback to manually writing zeroes. If + * fallocate() was used, unaligned reads from the fallocated + * area in raw_probe_alignment() will succeed, hence we need to + * allocate the first block. + * + * Optimize future alignment probing; ignore failures. + */ + allocate_first_block(fd, offset); } } else { result = 0; @@ -XXX,XX +XXX,XX @@ static int handle_aiocb_truncate(void *opaque) if (ftruncate(fd, offset) != 0) { result = -errno; error_setg_errno(errp, -result, "Could not resize file"); + } else if (current_length == 0 && offset > current_length) { + /* Optimize future alignment probing; ignore failures. */ + allocate_first_block(fd, offset); } return result; default: diff --git a/tests/qemu-iotests/059.out b/tests/qemu-iotests/059.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/059.out +++ b/tests/qemu-iotests/059.out @@ -XXX,XX +XXX,XX @@ Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824000 subformat=twoGbMax image: TEST_DIR/t.vmdk file format: vmdk virtual size: 0.977 TiB (1073741824000 bytes) -disk size: 16 KiB +disk size: 1.97 MiB Format specific information: cid: XXXXXXXX parent cid: XXXXXXXX diff --git a/tests/qemu-iotests/150.out b/tests/qemu-iotests/150.out.qcow2 similarity index 100% rename from tests/qemu-iotests/150.out rename to tests/qemu-iotests/150.out.qcow2 diff --git a/tests/qemu-iotests/150.out.raw b/tests/qemu-iotests/150.out.raw new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/tests/qemu-iotests/150.out.raw @@ -XXX,XX +XXX,XX @@ +QA output created by 150 + +=== Mapping sparse conversion === + +Offset Length File +0 0x1000 TEST_DIR/t.IMGFMT + +=== Mapping non-sparse conversion === + +Offset Length File +0 0x100000 TEST_DIR/t.IMGFMT +*** done diff --git a/tests/qemu-iotests/175 b/tests/qemu-iotests/175 index XXXXXXX..XXXXXXX 100755 --- a/tests/qemu-iotests/175 +++ b/tests/qemu-iotests/175 @@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15 # the file size. This function hides the resulting difference in the # stat -c '%b' output. # Parameter 1: Number of blocks an empty file occupies -# Parameter 2: Image size in bytes +# Parameter 2: Minimal number of blocks in an image +# Parameter 3: Image size in bytes _filter_blocks() { extra_blocks=$1 - img_size=$2 + min_blocks=$2 + img_size=$3 - sed -e "s/blocks=$extra_blocks\\(\$\\|[^0-9]\\)/nothing allocated/" \ - -e "s/blocks=$((extra_blocks + img_size / 512))\\(\$\\|[^0-9]\\)/everything allocated/" + sed -e "s/blocks=$min_blocks\\(\$\\|[^0-9]\\)/min allocation/" \ + -e "s/blocks=$((extra_blocks + img_size / 512))\\(\$\\|[^0-9]\\)/max allocation/" } # get standard environment, filters and checks @@ -XXX,XX +XXX,XX @@ size=$((1 * 1024 * 1024)) touch "$TEST_DIR/empty" extra_blocks=$(stat -c '%b' "$TEST_DIR/empty") +# We always write the first byte; check how many blocks this filesystem +# allocates to match empty image alloation. +printf "\0" > "$TEST_DIR/empty" +min_blocks=$(stat -c '%b' "$TEST_DIR/empty") + echo echo "== creating image with default preallocation ==" _make_test_img $size | _filter_imgfmt -stat -c "size=%s, blocks=%b" $TEST_IMG | _filter_blocks $extra_blocks $size +stat -c "size=%s, blocks=%b" $TEST_IMG | _filter_blocks $extra_blocks $min_blocks $size for mode in off full falloc; do echo echo "== creating image with preallocation $mode ==" IMGOPTS=preallocation=$mode _make_test_img $size | _filter_imgfmt - stat -c "size=%s, blocks=%b" $TEST_IMG | _filter_blocks $extra_blocks $size + stat -c "size=%s, blocks=%b" $TEST_IMG | _filter_blocks $extra_blocks $min_blocks $size done # success, all done diff --git a/tests/qemu-iotests/175.out b/tests/qemu-iotests/175.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/175.out +++ b/tests/qemu-iotests/175.out @@ -XXX,XX +XXX,XX @@ QA output created by 175 == creating image with default preallocation == Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1048576 -size=1048576, nothing allocated +size=1048576, min allocation == creating image with preallocation off == Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1048576 preallocation=off -size=1048576, nothing allocated +size=1048576, min allocation == creating image with preallocation full == Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1048576 preallocation=full -size=1048576, everything allocated +size=1048576, max allocation == creating image with preallocation falloc == Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1048576 preallocation=falloc -size=1048576, everything allocated +size=1048576, max allocation *** done diff --git a/tests/qemu-iotests/178.out.qcow2 b/tests/qemu-iotests/178.out.qcow2 index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/178.out.qcow2 +++ b/tests/qemu-iotests/178.out.qcow2 @@ -XXX,XX +XXX,XX @@ converted image file size in bytes: 196608 == raw input image with data (human) == Formatting 'TEST_DIR/t.qcow2', fmt=IMGFMT size=1073741824 -required size: 393216 +required size: 458752 fully allocated size: 1074135040 wrote 512/512 bytes at offset 512 512 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) @@ -XXX,XX +XXX,XX @@ converted image file size in bytes: 196608 Formatting 'TEST_DIR/t.qcow2', fmt=IMGFMT size=1073741824 { - "required": 393216, + "required": 458752, "fully-allocated": 1074135040 } wrote 512/512 bytes at offset 512 diff --git a/tests/qemu-iotests/221.out b/tests/qemu-iotests/221.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/221.out +++ b/tests/qemu-iotests/221.out @@ -XXX,XX +XXX,XX @@ QA output created by 221 === Check mapping of unaligned raw image === Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=65537 -[{ "start": 0, "length": 66048, "depth": 0, "zero": true, "data": false, "offset": OFFSET}] -[{ "start": 0, "length": 66048, "depth": 0, "zero": true, "data": false, "offset": OFFSET}] +[{ "start": 0, "length": 4096, "depth": 0, "zero": false, "data": true, "offset": OFFSET}, +{ "start": 4096, "length": 61952, "depth": 0, "zero": true, "data": false, "offset": OFFSET}] +[{ "start": 0, "length": 4096, "depth": 0, "zero": false, "data": true, "offset": OFFSET}, +{ "start": 4096, "length": 61952, "depth": 0, "zero": true, "data": false, "offset": OFFSET}] wrote 1/1 bytes at offset 65536 1 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) -[{ "start": 0, "length": 65536, "depth": 0, "zero": true, "data": false, "offset": OFFSET}, +[{ "start": 0, "length": 4096, "depth": 0, "zero": false, "data": true, "offset": OFFSET}, +{ "start": 4096, "length": 61440, "depth": 0, "zero": true, "data": false, "offset": OFFSET}, { "start": 65536, "length": 1, "depth": 0, "zero": false, "data": true, "offset": OFFSET}, { "start": 65537, "length": 511, "depth": 0, "zero": true, "data": false, "offset": OFFSET}] -[{ "start": 0, "length": 65536, "depth": 0, "zero": true, "data": false, "offset": OFFSET}, +[{ "start": 0, "length": 4096, "depth": 0, "zero": false, "data": true, "offset": OFFSET}, +{ "start": 4096, "length": 61440, "depth": 0, "zero": true, "data": false, "offset": OFFSET}, { "start": 65536, "length": 1, "depth": 0, "zero": false, "data": true, "offset": OFFSET}, { "start": 65537, "length": 511, "depth": 0, "zero": true, "data": false, "offset": OFFSET}] *** done diff --git a/tests/qemu-iotests/253.out b/tests/qemu-iotests/253.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/253.out +++ b/tests/qemu-iotests/253.out @@ -XXX,XX +XXX,XX @@ QA output created by 253 === Check mapping of unaligned raw image === Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1048575 -[{ "start": 0, "length": 1048576, "depth": 0, "zero": true, "data": false, "offset": OFFSET}] -[{ "start": 0, "length": 1048576, "depth": 0, "zero": true, "data": false, "offset": OFFSET}] +[{ "start": 0, "length": 4096, "depth": 0, "zero": false, "data": true, "offset": OFFSET}, +{ "start": 4096, "length": 1044480, "depth": 0, "zero": true, "data": false, "offset": OFFSET}] +[{ "start": 0, "length": 4096, "depth": 0, "zero": false, "data": true, "offset": OFFSET}, +{ "start": 4096, "length": 1044480, "depth": 0, "zero": true, "data": false, "offset": OFFSET}] wrote 65535/65535 bytes at offset 983040 63.999 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) -[{ "start": 0, "length": 983040, "depth": 0, "zero": true, "data": false, "offset": OFFSET}, +[{ "start": 0, "length": 4096, "depth": 0, "zero": false, "data": true, "offset": OFFSET}, +{ "start": 4096, "length": 978944, "depth": 0, "zero": true, "data": false, "offset": OFFSET}, { "start": 983040, "length": 65536, "depth": 0, "zero": false, "data": true, "offset": OFFSET}] -[{ "start": 0, "length": 983040, "depth": 0, "zero": true, "data": false, "offset": OFFSET}, +[{ "start": 0, "length": 4096, "depth": 0, "zero": false, "data": true, "offset": OFFSET}, +{ "start": 4096, "length": 978944, "depth": 0, "zero": true, "data": false, "offset": OFFSET}, { "start": 983040, "length": 65536, "depth": 0, "zero": false, "data": true, "offset": OFFSET}] *** done -- 2.21.0 From: Nir Soffer <nirsof@gmail.com> Using block_resize we can test allocate_first_block() with file descriptor opened with O_DIRECT, ensuring that it works for any size larger than 4096 bytes. Testing smaller sizes is tricky as the result depends on the filesystem used for testing. For example on NFS any size will work since O_DIRECT does not require any alignment. Signed-off-by: Nir Soffer <nsoffer@redhat.com> Reviewed-by: Max Reitz <mreitz@redhat.com> Message-id: 20190827010528.8818-3-nsoffer@redhat.com Signed-off-by: Max Reitz <mreitz@redhat.com> --- tests/qemu-iotests/175 | 28 ++++++++++++++++++++++++++++ tests/qemu-iotests/175.out | 8 ++++++++ 2 files changed, 36 insertions(+) diff --git a/tests/qemu-iotests/175 b/tests/qemu-iotests/175 index XXXXXXX..XXXXXXX 100755 --- a/tests/qemu-iotests/175 +++ b/tests/qemu-iotests/175 @@ -XXX,XX +XXX,XX @@ _filter_blocks() -e "s/blocks=$((extra_blocks + img_size / 512))\\(\$\\|[^0-9]\\)/max allocation/" } +# Resize image using block_resize. +# Parameter 1: image path +# Parameter 2: new size +_block_resize() +{ + local path=$1 + local size=$2 + + $QEMU -qmp stdio -nographic -nodefaults \ + -blockdev file,node-name=file,filename=$path,cache.direct=on \ + <<EOF +{'execute': 'qmp_capabilities'} +{'execute': 'block_resize', 'arguments': {'node-name': 'file', 'size': $size}} +{'execute': 'quit'} +EOF +} + # get standard environment, filters and checks . ./common.rc . ./common.filter @@ -XXX,XX +XXX,XX @@ _supported_fmt raw _supported_proto file _supported_os Linux +_default_cache_mode none +_supported_cache_modes none directsync + size=$((1 * 1024 * 1024)) touch "$TEST_DIR/empty" @@ -XXX,XX +XXX,XX @@ for mode in off full falloc; do stat -c "size=%s, blocks=%b" $TEST_IMG | _filter_blocks $extra_blocks $min_blocks $size done +for new_size in 4096 1048576; do + echo + echo "== resize empty image with block_resize ==" + _make_test_img 0 | _filter_imgfmt + _block_resize $TEST_IMG $new_size >/dev/null + stat -c "size=%s, blocks=%b" $TEST_IMG | _filter_blocks $extra_blocks $min_blocks $new_size +done + # success, all done echo "*** done" rm -f $seq.full diff --git a/tests/qemu-iotests/175.out b/tests/qemu-iotests/175.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/175.out +++ b/tests/qemu-iotests/175.out @@ -XXX,XX +XXX,XX @@ size=1048576, max allocation == creating image with preallocation falloc == Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1048576 preallocation=falloc size=1048576, max allocation + +== resize empty image with block_resize == +Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=0 +size=4096, min allocation + +== resize empty image with block_resize == +Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=0 +size=1048576, min allocation *** done -- 2.21.0 fe646693acc changed qemu-img create's output so that it no longer prints single quotes around parameter values. The subformat and adapter_type filters in _filter_img_create() have never been adapted to that change. Fixes: fe646693acc13ac48b98435d14149ab04dc597bc Signed-off-by: Max Reitz <mreitz@redhat.com> Reviewed-by: John Snow <jsnow@redhat.com> Message-id: 20190815153638.4600-2-mreitz@redhat.com Reviewed-by: John Snow <jsnow@redhat.com> Signed-off-by: Max Reitz <mreitz@redhat.com> --- tests/qemu-iotests/059.out | 16 ++++++++-------- tests/qemu-iotests/common.filter | 4 ++-- 2 files changed, 10 insertions(+), 10 deletions(-) diff --git a/tests/qemu-iotests/059.out b/tests/qemu-iotests/059.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/059.out +++ b/tests/qemu-iotests/059.out @@ -XXX,XX +XXX,XX @@ Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864 qemu-io: can't open device TEST_DIR/t.vmdk: L1 size too big === Testing monolithicFlat creation and opening === -Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=2147483648 subformat=monolithicFlat +Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=2147483648 image: TEST_DIR/t.IMGFMT file format: IMGFMT virtual size: 2 GiB (2147483648 bytes) === Testing monolithicFlat with zeroed_grain === qemu-img: TEST_DIR/t.IMGFMT: Flat image can't enable zeroed grain -Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=2147483648 subformat=monolithicFlat +Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=2147483648 === Testing big twoGbMaxExtentFlat === -Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824000 subformat=twoGbMaxExtentFlat +Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824000 image: TEST_DIR/t.vmdk file format: vmdk virtual size: 0.977 TiB (1073741824000 bytes) @@ -XXX,XX +XXX,XX @@ Format specific information: qemu-img: Could not open 'TEST_DIR/t.IMGFMT': Invalid extent line: RW 12582912 VMFS "dummy.IMGFMT" 1 === Testing truncated sparse === -Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=107374182400 subformat=monolithicSparse +Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=107374182400 qemu-img: Could not open 'TEST_DIR/t.IMGFMT': File truncated, expecting at least 13172736 bytes === Converting to streamOptimized from image with small cluster size=== @@ -XXX,XX +XXX,XX @@ wrote 512/512 bytes at offset 10240 512 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) === Testing monolithicFlat with internally generated JSON file name === -Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864 subformat=monolithicFlat +Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864 qemu-io: can't open: Cannot use relative extent paths with VMDK descriptor file 'json:{"image": {"driver": "file", "filename": "TEST_DIR/t.IMGFMT"}, "driver": "blkdebug", "inject-error.0.event": "read_aio"}' === Testing version 3 === @@ -XXX,XX +XXX,XX @@ read 512/512 bytes at offset 64931328 512 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) === Testing 4TB monolithicFlat creation and IO === -Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=4398046511104 subformat=monolithicFlat +Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=4398046511104 image: TEST_DIR/t.IMGFMT file format: IMGFMT virtual size: 4 TiB (4398046511104 bytes) @@ -XXX,XX +XXX,XX @@ read 1024/1024 bytes at offset 966367641600 1 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) === Testing qemu-img map on extents === -Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=33285996544 subformat=monolithicSparse +Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=33285996544 wrote 1024/1024 bytes at offset 65024 1 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) wrote 1024/1024 bytes at offset 2147483136 @@ -XXX,XX +XXX,XX @@ Offset Length Mapped to File 0 0x20000 0x3f0000 TEST_DIR/t.vmdk 0x7fff0000 0x20000 0x410000 TEST_DIR/t.vmdk 0x140000000 0x10000 0x430000 TEST_DIR/t.vmdk -Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=33285996544 subformat=twoGbMaxExtentSparse +Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=33285996544 wrote 1024/1024 bytes at offset 65024 1 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) wrote 1024/1024 bytes at offset 2147483136 diff --git a/tests/qemu-iotests/common.filter b/tests/qemu-iotests/common.filter index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/common.filter +++ b/tests/qemu-iotests/common.filter @@ -XXX,XX +XXX,XX @@ _filter_img_create() -e "s# compat6=\\(on\\|off\\)##g" \ -e "s# static=\\(on\\|off\\)##g" \ -e "s# zeroed_grain=\\(on\\|off\\)##g" \ - -e "s# subformat='[^']*'##g" \ - -e "s# adapter_type='[^']*'##g" \ + -e "s# subformat=[^ ]*##g" \ + -e "s# adapter_type=[^ ]*##g" \ -e "s# hwversion=[^ ]*##g" \ -e "s# lazy_refcounts=\\(on\\|off\\)##g" \ -e "s# block_size=[0-9]\\+##g" \ -- 2.21.0 This makes iotest 033 pass with e.g. subformat=monolithicFlat. It also turns a former error in 059 into success. Signed-off-by: Max Reitz <mreitz@redhat.com> Message-id: 20190815153638.4600-3-mreitz@redhat.com Reviewed-by: John Snow <jsnow@redhat.com> Signed-off-by: Max Reitz <mreitz@redhat.com> --- block/vmdk.c | 54 ++++++++++++++++++++++++-------------- tests/qemu-iotests/059 | 7 +++-- tests/qemu-iotests/059.out | 4 ++- 3 files changed, 42 insertions(+), 23 deletions(-) diff --git a/block/vmdk.c b/block/vmdk.c index XXXXXXX..XXXXXXX 100644 --- a/block/vmdk.c +++ b/block/vmdk.c @@ -XXX,XX +XXX,XX @@ static const char *next_line(const char *s) } static int vmdk_parse_extents(const char *desc, BlockDriverState *bs, - const char *desc_file_path, QDict *options, - Error **errp) + QDict *options, Error **errp) { int ret; int matches; @@ -XXX,XX +XXX,XX @@ static int vmdk_parse_extents(const char *desc, BlockDriverState *bs, const char *p, *np; int64_t sectors = 0; int64_t flat_offset; + char *desc_file_dir = NULL; char *extent_path; BdrvChild *extent_file; BDRVVmdkState *s = bs->opaque; @@ -XXX,XX +XXX,XX @@ static int vmdk_parse_extents(const char *desc, BlockDriverState *bs, continue; } - if (!path_is_absolute(fname) && !path_has_protocol(fname) && - !desc_file_path[0]) - { - bdrv_refresh_filename(bs->file->bs); - error_setg(errp, "Cannot use relative extent paths with VMDK " - "descriptor file '%s'", bs->file->bs->filename); - return -EINVAL; - } + if (path_is_absolute(fname)) { + extent_path = g_strdup(fname); + } else { + if (!desc_file_dir) { + desc_file_dir = bdrv_dirname(bs->file->bs, errp); + if (!desc_file_dir) { + bdrv_refresh_filename(bs->file->bs); + error_prepend(errp, "Cannot use relative paths with VMDK " + "descriptor file '%s': ", + bs->file->bs->filename); + ret = -EINVAL; + goto out; + } + } - extent_path = path_combine(desc_file_path, fname); + extent_path = g_strconcat(desc_file_dir, fname, NULL); + } ret = snprintf(extent_opt_prefix, 32, "extents.%d", s->num_extents); assert(ret < 32); @@ -XXX,XX +XXX,XX @@ static int vmdk_parse_extents(const char *desc, BlockDriverState *bs, g_free(extent_path); if (local_err) { error_propagate(errp, local_err); - return -EINVAL; + ret = -EINVAL; + goto out; } /* save to extents array */ @@ -XXX,XX +XXX,XX @@ static int vmdk_parse_extents(const char *desc, BlockDriverState *bs, 0, 0, 0, 0, 0, &extent, errp); if (ret < 0) { bdrv_unref_child(bs, extent_file); - return ret; + goto out; } extent->flat_start_offset = flat_offset << 9; } else if (!strcmp(type, "SPARSE") || !strcmp(type, "VMFSSPARSE")) { @@ -XXX,XX +XXX,XX @@ static int vmdk_parse_extents(const char *desc, BlockDriverState *bs, g_free(buf); if (ret) { bdrv_unref_child(bs, extent_file); - return ret; + goto out; } extent = &s->extents[s->num_extents - 1]; } else if (!strcmp(type, "SESPARSE")) { ret = vmdk_open_se_sparse(bs, extent_file, bs->open_flags, errp); if (ret) { bdrv_unref_child(bs, extent_file); - return ret; + goto out; } extent = &s->extents[s->num_extents - 1]; } else { error_setg(errp, "Unsupported extent type '%s'", type); bdrv_unref_child(bs, extent_file); - return -ENOTSUP; + ret = -ENOTSUP; + goto out; } extent->type = g_strdup(type); } - return 0; + + ret = 0; + goto out; invalid: np = next_line(p); @@ -XXX,XX +XXX,XX @@ invalid: np--; } error_setg(errp, "Invalid extent line: %.*s", (int)(np - p), p); - return -EINVAL; + ret = -EINVAL; + +out: + g_free(desc_file_dir); + return ret; } static int vmdk_open_desc_file(BlockDriverState *bs, int flags, char *buf, @@ -XXX,XX +XXX,XX @@ static int vmdk_open_desc_file(BlockDriverState *bs, int flags, char *buf, } s->create_type = g_strdup(ct); s->desc_offset = 0; - ret = vmdk_parse_extents(buf, bs, bs->file->bs->exact_filename, options, - errp); + ret = vmdk_parse_extents(buf, bs, options, errp); exit: return ret; } diff --git a/tests/qemu-iotests/059 b/tests/qemu-iotests/059 index XXXXXXX..XXXXXXX 100755 --- a/tests/qemu-iotests/059 +++ b/tests/qemu-iotests/059 @@ -XXX,XX +XXX,XX @@ $QEMU_IMG convert -f qcow2 -O vmdk -o subformat=streamOptimized "$TEST_IMG.qcow2 echo echo "=== Testing monolithicFlat with internally generated JSON file name ===" +# Should work, because bdrv_dirname() works fine with blkdebug IMGOPTS="subformat=monolithicFlat" _make_test_img 64M -$QEMU_IO -c "open -o driver=$IMGFMT,file.driver=blkdebug,file.image.filename=$TEST_IMG,file.inject-error.0.event=read_aio" 2>&1 \ - | _filter_testdir | _filter_imgfmt +$QEMU_IO -c "open -o driver=$IMGFMT,file.driver=blkdebug,file.image.filename=$TEST_IMG,file.inject-error.0.event=read_aio" \ + -c info \ + 2>&1 \ + | _filter_testdir | _filter_imgfmt | _filter_img_info _cleanup_test_img echo diff --git a/tests/qemu-iotests/059.out b/tests/qemu-iotests/059.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/059.out +++ b/tests/qemu-iotests/059.out @@ -XXX,XX +XXX,XX @@ wrote 512/512 bytes at offset 10240 === Testing monolithicFlat with internally generated JSON file name === Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864 -qemu-io: can't open: Cannot use relative extent paths with VMDK descriptor file 'json:{"image": {"driver": "file", "filename": "TEST_DIR/t.IMGFMT"}, "driver": "blkdebug", "inject-error.0.event": "read_aio"}' +format name: IMGFMT +cluster size: 0 bytes +vm state offset: 0 bytes === Testing version 3 === image: TEST_DIR/iotest-version3.IMGFMT -- 2.21.0 We had a test for a case where relative extent paths did not work, but unfortunately we just fixed the underlying problem, so it works now. This patch adds a new test case that still fails. Signed-off-by: Max Reitz <mreitz@redhat.com> Reviewed-by: John Snow <jsnow@redhat.com> Message-id: 20190815153638.4600-4-mreitz@redhat.com Reviewed-by: John Snow <jsnow@redhat.com> Signed-off-by: Max Reitz <mreitz@redhat.com> --- tests/qemu-iotests/059 | 27 +++++++++++++++++++++++++++ tests/qemu-iotests/059.out | 4 ++++ 2 files changed, 31 insertions(+) diff --git a/tests/qemu-iotests/059 b/tests/qemu-iotests/059 index XXXXXXX..XXXXXXX 100755 --- a/tests/qemu-iotests/059 +++ b/tests/qemu-iotests/059 @@ -XXX,XX +XXX,XX @@ $QEMU_IMG convert -f qcow2 -O vmdk -o subformat=streamOptimized "$TEST_IMG.qcow2 echo echo "=== Testing monolithicFlat with internally generated JSON file name ===" + +echo '--- blkdebug ---' # Should work, because bdrv_dirname() works fine with blkdebug IMGOPTS="subformat=monolithicFlat" _make_test_img 64M $QEMU_IO -c "open -o driver=$IMGFMT,file.driver=blkdebug,file.image.filename=$TEST_IMG,file.inject-error.0.event=read_aio" \ @@ -XXX,XX +XXX,XX @@ $QEMU_IO -c "open -o driver=$IMGFMT,file.driver=blkdebug,file.image.filename=$TE | _filter_testdir | _filter_imgfmt | _filter_img_info _cleanup_test_img +echo '--- quorum ---' +# Should not work, because bdrv_dirname() does not work with quorum +IMGOPTS="subformat=monolithicFlat" _make_test_img 64M +cp "$TEST_IMG" "$TEST_IMG.orig" + +filename="json:{ + \"driver\": \"$IMGFMT\", + \"file\": { + \"driver\": \"quorum\", + \"children\": [ { + \"driver\": \"file\", + \"filename\": \"$TEST_IMG\" + }, { + \"driver\": \"file\", + \"filename\": \"$TEST_IMG.orig\" + } ], + \"vote-threshold\": 1 + } }" + +filename=$(echo "$filename" | tr '\n' ' ' | sed -e 's/\s\+/ /g') +$QEMU_IMG info "$filename" 2>&1 \ + | sed -e "s/'json:[^']*'/\$QUORUM_FILE/g" \ + | _filter_testdir | _filter_imgfmt | _filter_img_info + + echo echo "=== Testing version 3 ===" _use_sample_img iotest-version3.vmdk.bz2 diff --git a/tests/qemu-iotests/059.out b/tests/qemu-iotests/059.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/059.out +++ b/tests/qemu-iotests/059.out @@ -XXX,XX +XXX,XX @@ wrote 512/512 bytes at offset 10240 512 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) === Testing monolithicFlat with internally generated JSON file name === +--- blkdebug --- Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864 format name: IMGFMT cluster size: 0 bytes vm state offset: 0 bytes +--- quorum --- +Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864 +qemu-img: Could not open $QUORUM_FILE: Cannot use relative paths with VMDK descriptor file $QUORUM_FILE: Cannot generate a base directory for quorum nodes === Testing version 3 === image: TEST_DIR/iotest-version3.IMGFMT -- 2.21.0 Compressed writes generally have to write full clusters, not just in theory but also in practice when it comes to vmdk's streamOptimized subformat. It currently is just silently broken for writes with non-zero in-cluster offsets: $ qemu-img create -f vmdk -o subformat=streamOptimized foo.vmdk 1M $ qemu-io -c 'write 4k 4k' -c 'read 4k 4k' foo.vmdk wrote 4096/4096 bytes at offset 4096 4 KiB, 1 ops; 00.01 sec (443.724 KiB/sec and 110.9309 ops/sec) read failed: Invalid argument (The technical reason is that vmdk_write_extent() just writes the incomplete compressed data actually to offset 4k. When reading the data, vmdk_read_extent() looks at offset 0 and finds the compressed data size to be 0, because that is what it reads from there. This yields an error.) For incomplete writes with zero in-cluster offsets, the error path when reading the rest of the cluster is a bit different, but the result is the same: $ qemu-img create -f vmdk -o subformat=streamOptimized foo.vmdk 1M $ qemu-io -c 'write 0k 4k' -c 'read 4k 4k' foo.vmdk wrote 4096/4096 bytes at offset 0 4 KiB, 1 ops; 00.01 sec (362.641 KiB/sec and 90.6603 ops/sec) read failed: Invalid argument (Here, vmdk_read_extent() finds the data and then sees that the uncompressed data is short.) It is better to reject invalid writes than to make the user believe they might have succeeded and then fail when trying to read it back. Signed-off-by: Max Reitz <mreitz@redhat.com> Reviewed-by: John Snow <jsnow@redhat.com> Message-id: 20190815153638.4600-5-mreitz@redhat.com Reviewed-by: John Snow <jsnow@redhat.com> Signed-off-by: Max Reitz <mreitz@redhat.com> --- block/vmdk.c | 10 ++++++++++ 1 file changed, 10 insertions(+) diff --git a/block/vmdk.c b/block/vmdk.c index XXXXXXX..XXXXXXX 100644 --- a/block/vmdk.c +++ b/block/vmdk.c @@ -XXX,XX +XXX,XX @@ static int vmdk_write_extent(VmdkExtent *extent, int64_t cluster_offset, if (extent->compressed) { void *compressed_data; + /* Only whole clusters */ + if (offset_in_cluster || + n_bytes > (extent->cluster_sectors * SECTOR_SIZE) || + (n_bytes < (extent->cluster_sectors * SECTOR_SIZE) && + offset + n_bytes != extent->end_sector * SECTOR_SIZE)) + { + ret = -EINVAL; + goto out; + } + if (!extent->has_marker) { ret = -EINVAL; goto out; -- 2.21.0 streamOptimized does not support writes that do not span exactly one cluster. Furthermore, it cannot rewrite already allocated clusters. As such, many iotests do not work with it. Disable them. Signed-off-by: Max Reitz <mreitz@redhat.com> Message-id: 20190815153638.4600-6-mreitz@redhat.com Reviewed-by: John Snow <jsnow@redhat.com> Signed-off-by: Max Reitz <mreitz@redhat.com> --- tests/qemu-iotests/002 | 1 + tests/qemu-iotests/003 | 1 + tests/qemu-iotests/005 | 3 ++- tests/qemu-iotests/009 | 1 + tests/qemu-iotests/010 | 1 + tests/qemu-iotests/011 | 1 + tests/qemu-iotests/017 | 3 ++- tests/qemu-iotests/018 | 3 ++- tests/qemu-iotests/019 | 3 ++- tests/qemu-iotests/020 | 3 ++- tests/qemu-iotests/027 | 1 + tests/qemu-iotests/032 | 1 + tests/qemu-iotests/033 | 1 + tests/qemu-iotests/034 | 3 ++- tests/qemu-iotests/037 | 3 ++- tests/qemu-iotests/063 | 3 ++- tests/qemu-iotests/072 | 1 + tests/qemu-iotests/105 | 3 ++- tests/qemu-iotests/197 | 1 + tests/qemu-iotests/215 | 1 + tests/qemu-iotests/251 | 1 + 21 files changed, 30 insertions(+), 9 deletions(-) diff --git a/tests/qemu-iotests/002 b/tests/qemu-iotests/002 index XXXXXXX..XXXXXXX 100755 --- a/tests/qemu-iotests/002 +++ b/tests/qemu-iotests/002 @@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15 _supported_fmt generic _supported_proto generic +_unsupported_imgopts "subformat=streamOptimized" size=128M diff --git a/tests/qemu-iotests/003 b/tests/qemu-iotests/003 index XXXXXXX..XXXXXXX 100755 --- a/tests/qemu-iotests/003 +++ b/tests/qemu-iotests/003 @@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15 _supported_fmt generic _supported_proto generic +_unsupported_imgopts "subformat=streamOptimized" size=128M offset=67M diff --git a/tests/qemu-iotests/005 b/tests/qemu-iotests/005 index XXXXXXX..XXXXXXX 100755 --- a/tests/qemu-iotests/005 +++ b/tests/qemu-iotests/005 @@ -XXX,XX +XXX,XX @@ _supported_fmt generic _supported_proto generic _supported_os Linux _unsupported_imgopts "subformat=twoGbMaxExtentFlat" \ - "subformat=twoGbMaxExtentSparse" + "subformat=twoGbMaxExtentSparse" \ + "subformat=streamOptimized" # vpc is limited to 127GB, so we can't test it here if [ "$IMGFMT" = "vpc" ]; then diff --git a/tests/qemu-iotests/009 b/tests/qemu-iotests/009 index XXXXXXX..XXXXXXX 100755 --- a/tests/qemu-iotests/009 +++ b/tests/qemu-iotests/009 @@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15 _supported_fmt generic _supported_proto generic +_unsupported_imgopts "subformat=streamOptimized" size=6G diff --git a/tests/qemu-iotests/010 b/tests/qemu-iotests/010 index XXXXXXX..XXXXXXX 100755 --- a/tests/qemu-iotests/010 +++ b/tests/qemu-iotests/010 @@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15 _supported_fmt generic _supported_proto generic +_unsupported_imgopts "subformat=streamOptimized" size=6G diff --git a/tests/qemu-iotests/011 b/tests/qemu-iotests/011 index XXXXXXX..XXXXXXX 100755 --- a/tests/qemu-iotests/011 +++ b/tests/qemu-iotests/011 @@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15 _supported_fmt generic _supported_proto generic +_unsupported_imgopts "subformat=streamOptimized" size=6G diff --git a/tests/qemu-iotests/017 b/tests/qemu-iotests/017 index XXXXXXX..XXXXXXX 100755 --- a/tests/qemu-iotests/017 +++ b/tests/qemu-iotests/017 @@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15 _supported_fmt qcow qcow2 vmdk qed _supported_proto generic _unsupported_proto vxhs -_unsupported_imgopts "subformat=monolithicFlat" "subformat=twoGbMaxExtentFlat" +_unsupported_imgopts "subformat=monolithicFlat" "subformat=twoGbMaxExtentFlat" \ + "subformat=streamOptimized" TEST_OFFSETS="0 4294967296" diff --git a/tests/qemu-iotests/018 b/tests/qemu-iotests/018 index XXXXXXX..XXXXXXX 100755 --- a/tests/qemu-iotests/018 +++ b/tests/qemu-iotests/018 @@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15 _supported_fmt qcow qcow2 vmdk qed _supported_proto file _supported_os Linux -_unsupported_imgopts "subformat=monolithicFlat" "subformat=twoGbMaxExtentFlat" +_unsupported_imgopts "subformat=monolithicFlat" "subformat=twoGbMaxExtentFlat" \ + "streamOptimized" TEST_OFFSETS="0 4294967296" diff --git a/tests/qemu-iotests/019 b/tests/qemu-iotests/019 index XXXXXXX..XXXXXXX 100755 --- a/tests/qemu-iotests/019 +++ b/tests/qemu-iotests/019 @@ -XXX,XX +XXX,XX @@ _supported_proto file _supported_os Linux _unsupported_imgopts "subformat=monolithicFlat" \ "subformat=twoGbMaxExtentFlat" \ - "subformat=twoGbMaxExtentSparse" + "subformat=twoGbMaxExtentSparse" \ + "subformat=streamOptimized" TEST_OFFSETS="0 4294967296" CLUSTER_SIZE=65536 diff --git a/tests/qemu-iotests/020 b/tests/qemu-iotests/020 index XXXXXXX..XXXXXXX 100755 --- a/tests/qemu-iotests/020 +++ b/tests/qemu-iotests/020 @@ -XXX,XX +XXX,XX @@ _supported_fmt qcow qcow2 vmdk qed _supported_proto file _unsupported_imgopts "subformat=monolithicFlat" \ "subformat=twoGbMaxExtentFlat" \ - "subformat=twoGbMaxExtentSparse" + "subformat=twoGbMaxExtentSparse" \ + "subformat=streamOptimized" TEST_OFFSETS="0 4294967296" diff --git a/tests/qemu-iotests/027 b/tests/qemu-iotests/027 index XXXXXXX..XXXXXXX 100755 --- a/tests/qemu-iotests/027 +++ b/tests/qemu-iotests/027 @@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15 _supported_fmt vmdk qcow qcow2 qed _supported_proto generic +_unsupported_imgopts "subformat=streamOptimized" size=128M diff --git a/tests/qemu-iotests/032 b/tests/qemu-iotests/032 index XXXXXXX..XXXXXXX 100755 --- a/tests/qemu-iotests/032 +++ b/tests/qemu-iotests/032 @@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15 # This works for any image format (though unlikely to segfault for raw) _supported_fmt generic _supported_proto generic +_unsupported_imgopts "subformat=streamOptimized" echo echo === Prepare image === diff --git a/tests/qemu-iotests/033 b/tests/qemu-iotests/033 index XXXXXXX..XXXXXXX 100755 --- a/tests/qemu-iotests/033 +++ b/tests/qemu-iotests/033 @@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15 _supported_fmt generic _supported_proto generic +_unsupported_imgopts "subformat=streamOptimized" size=128M diff --git a/tests/qemu-iotests/034 b/tests/qemu-iotests/034 index XXXXXXX..XXXXXXX 100755 --- a/tests/qemu-iotests/034 +++ b/tests/qemu-iotests/034 @@ -XXX,XX +XXX,XX @@ _supported_proto file _supported_os Linux _unsupported_imgopts "subformat=monolithicFlat" \ "subformat=twoGbMaxExtentFlat" \ - "subformat=twoGbMaxExtentSparse" + "subformat=twoGbMaxExtentSparse" \ + "subformat=streamOptimized" CLUSTER_SIZE=4k size=128M diff --git a/tests/qemu-iotests/037 b/tests/qemu-iotests/037 index XXXXXXX..XXXXXXX 100755 --- a/tests/qemu-iotests/037 +++ b/tests/qemu-iotests/037 @@ -XXX,XX +XXX,XX @@ _supported_fmt qcow qcow2 vmdk qed _supported_proto file _unsupported_imgopts "subformat=monolithicFlat" \ "subformat=twoGbMaxExtentFlat" \ - "subformat=twoGbMaxExtentSparse" + "subformat=twoGbMaxExtentSparse" \ + "subformat=streamOptimized" CLUSTER_SIZE=4k size=128M diff --git a/tests/qemu-iotests/063 b/tests/qemu-iotests/063 index XXXXXXX..XXXXXXX 100755 --- a/tests/qemu-iotests/063 +++ b/tests/qemu-iotests/063 @@ -XXX,XX +XXX,XX @@ _supported_fmt qcow qcow2 vmdk qed raw _supported_proto file _unsupported_imgopts "subformat=monolithicFlat" \ "subformat=twoGbMaxExtentFlat" \ - "subformat=twoGbMaxExtentSparse" + "subformat=twoGbMaxExtentSparse" \ + "subformat=streamOptimized" _make_test_img 4M diff --git a/tests/qemu-iotests/072 b/tests/qemu-iotests/072 index XXXXXXX..XXXXXXX 100755 --- a/tests/qemu-iotests/072 +++ b/tests/qemu-iotests/072 @@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15 _supported_fmt vpc vmdk vhdx vdi qed qcow2 qcow _supported_proto file +_unsupported_imgopts "subformat=streamOptimized" IMG_SIZE=64M diff --git a/tests/qemu-iotests/105 b/tests/qemu-iotests/105 index XXXXXXX..XXXXXXX 100755 --- a/tests/qemu-iotests/105 +++ b/tests/qemu-iotests/105 @@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15 _supported_fmt qcow2 vmdk vhdx qed _supported_proto generic _unsupported_imgopts "subformat=twoGbMaxExtentFlat" \ - "subformat=twoGbMaxExtentSparse" + "subformat=twoGbMaxExtentSparse" \ + "subformat=streamOptimized" echo echo "creating large image" diff --git a/tests/qemu-iotests/197 b/tests/qemu-iotests/197 index XXXXXXX..XXXXXXX 100755 --- a/tests/qemu-iotests/197 +++ b/tests/qemu-iotests/197 @@ -XXX,XX +XXX,XX @@ _supported_fmt generic _supported_proto generic # LUKS support may be possible, but it complicates things. _unsupported_fmt luks +_unsupported_imgopts "subformat=streamOptimized" echo echo '=== Copy-on-read ===' diff --git a/tests/qemu-iotests/215 b/tests/qemu-iotests/215 index XXXXXXX..XXXXXXX 100755 --- a/tests/qemu-iotests/215 +++ b/tests/qemu-iotests/215 @@ -XXX,XX +XXX,XX @@ _supported_fmt generic _supported_proto generic # LUKS support may be possible, but it complicates things. _unsupported_fmt luks +_unsupported_imgopts "subformat=streamOptimized" echo echo '=== Copy-on-read ===' diff --git a/tests/qemu-iotests/251 b/tests/qemu-iotests/251 index XXXXXXX..XXXXXXX 100755 --- a/tests/qemu-iotests/251 +++ b/tests/qemu-iotests/251 @@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15 _supported_fmt generic _supported_proto file _supported_os Linux +_unsupported_imgopts "subformat=streamOptimized" if [ "$IMGOPTSSYNTAX" = "true" ]; then # We use json:{} filenames here, so we cannot work with additional options. -- 2.21.0 The error message for the test case where we have a quorum node for which no directory name can be generated is different: For twoGbMaxExtentSparse, it complains that it cannot open the extent file. For other (sub)formats, it just notes that it cannot determine the backing file path. Both are fine, but just disable twoGbMaxExtentSparse for simplicity's sake. Signed-off-by: Max Reitz <mreitz@redhat.com> Reviewed-by: John Snow <jsnow@redhat.com> Message-id: 20190815153638.4600-7-mreitz@redhat.com Reviewed-by: John Snow <jsnow@redhat.com> Signed-off-by: Max Reitz <mreitz@redhat.com> --- tests/qemu-iotests/110 | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/tests/qemu-iotests/110 b/tests/qemu-iotests/110 index XXXXXXX..XXXXXXX 100755 --- a/tests/qemu-iotests/110 +++ b/tests/qemu-iotests/110 @@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15 # Any format supporting backing files _supported_fmt qed qcow qcow2 vmdk _supported_proto file -_unsupported_imgopts "subformat=monolithicFlat" "subformat=twoGbMaxExtentFlat" +_unsupported_imgopts "subformat=monolithicFlat" "subformat=twoGbMaxExtentFlat" \ + "subformat=twoGbMaxExtentSparse" TEST_IMG_REL=$(basename "$TEST_IMG") -- 2.21.0 iotest 126 requires backing file support, which flat vmdks cannot offer. Skip this test for such subformats. Signed-off-by: Max Reitz <mreitz@redhat.com> Message-id: 20190815153638.4600-8-mreitz@redhat.com Reviewed-by: John Snow <jsnow@redhat.com> Signed-off-by: Max Reitz <mreitz@redhat.com> --- tests/qemu-iotests/126 | 2 ++ 1 file changed, 2 insertions(+) diff --git a/tests/qemu-iotests/126 b/tests/qemu-iotests/126 index XXXXXXX..XXXXXXX 100755 --- a/tests/qemu-iotests/126 +++ b/tests/qemu-iotests/126 @@ -XXX,XX +XXX,XX @@ status=1 # failure is the default! # Needs backing file support _supported_fmt qcow qcow2 qed vmdk +_unsupported_imgopts "subformat=monolithicFlat" \ + "subformat=twoGbMaxExtentFlat" # This is the default protocol (and we want to test the difference between # colons which separate a protocol prefix from the rest and colons which are # just part of the filename, so we cannot test protocols which require a prefix) -- 2.21.0 From: Stefan Hajnoczi <stefanha@redhat.com> Fixes: a6b257a08e3d72219f03e461a52152672fec0612 ("file-posix: Handle undetectable alignment") Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Message-id: 20190827101328.4062-1-stefanha@redhat.com Reviewed-by: Thomas Huth <thuth@redhat.com> Signed-off-by: Max Reitz <mreitz@redhat.com> --- block/file-posix.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/block/file-posix.c b/block/file-posix.c index XXXXXXX..XXXXXXX 100644 --- a/block/file-posix.c +++ b/block/file-posix.c @@ -XXX,XX +XXX,XX @@ static void raw_probe_alignment(BlockDriverState *bs, int fd, Error **errp) for (i = 0; i < ARRAY_SIZE(alignments); i++) { align = alignments[i]; if (raw_is_io_aligned(fd, buf + align, max_align)) { - /* Fallback to request_aligment. */ + /* Fallback to request_alignment. */ s->buf_align = (align != 1) ? align : bs->bl.request_alignment; break; } -- 2.21.0 From: Thomas Huth <thuth@redhat.com> It is possible to enable only a subset of the block drivers with the "--block-drv-rw-whitelist" option of the "configure" script. All other drivers are marked as unusable (or only included as read-only with the "--block-drv-ro-whitelist" option). If an iotest is now using such a disabled block driver, it is failing - which is bad, since at least the tests in the "auto" group should be able to deal with this situation. Thus let's introduce a "_require_drivers" function that can be used by the shell tests to check for the availability of certain drivers first, and marks the test as "not run" if one of the drivers is missing. This patch mainly targets the test in the "auto" group which should never fail in such a case, but also improves some of the other tests along the way. Note that we also assume that the "qcow2" and "file" drivers are always available - otherwise it does not make sense to run "make check-block" at all (which only tests with qcow2 by default). Signed-off-by: Thomas Huth <thuth@redhat.com> Message-id: 20190823133552.11680-1-thuth@redhat.com Signed-off-by: Max Reitz <mreitz@redhat.com> --- tests/qemu-iotests/071 | 1 + tests/qemu-iotests/081 | 4 +--- tests/qemu-iotests/099 | 1 + tests/qemu-iotests/120 | 1 + tests/qemu-iotests/162 | 4 +--- tests/qemu-iotests/184 | 1 + tests/qemu-iotests/186 | 1 + tests/qemu-iotests/common.rc | 14 ++++++++++++++ 8 files changed, 21 insertions(+), 6 deletions(-) diff --git a/tests/qemu-iotests/071 b/tests/qemu-iotests/071 index XXXXXXX..XXXXXXX 100755 --- a/tests/qemu-iotests/071 +++ b/tests/qemu-iotests/071 @@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15 _supported_fmt qcow2 _supported_proto file +_require_drivers blkdebug blkverify do_run_qemu() { diff --git a/tests/qemu-iotests/081 b/tests/qemu-iotests/081 index XXXXXXX..XXXXXXX 100755 --- a/tests/qemu-iotests/081 +++ b/tests/qemu-iotests/081 @@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15 _supported_fmt raw _supported_proto file _supported_os Linux +_require_drivers quorum do_run_qemu() { @@ -XXX,XX +XXX,XX @@ run_qemu() | _filter_qemu_io | _filter_generated_node_ids } -test_quorum=$($QEMU_IMG --help|grep quorum) -[ "$test_quorum" = "" ] && _supported_fmt quorum - quorum="driver=raw,file.driver=quorum,file.vote-threshold=2" quorum="$quorum,file.children.0.file.filename=$TEST_DIR/1.raw" quorum="$quorum,file.children.1.file.filename=$TEST_DIR/2.raw" diff --git a/tests/qemu-iotests/099 b/tests/qemu-iotests/099 index XXXXXXX..XXXXXXX 100755 --- a/tests/qemu-iotests/099 +++ b/tests/qemu-iotests/099 @@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15 _supported_fmt qcow qcow2 qed vdi vhdx vmdk vpc _supported_proto file _supported_os Linux +_require_drivers blkdebug blkverify _unsupported_imgopts "subformat=monolithicFlat" "subformat=twoGbMaxExtentFlat" \ "subformat=twoGbMaxExtentSparse" diff --git a/tests/qemu-iotests/120 b/tests/qemu-iotests/120 index XXXXXXX..XXXXXXX 100755 --- a/tests/qemu-iotests/120 +++ b/tests/qemu-iotests/120 @@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15 _supported_fmt generic _supported_proto file _unsupported_fmt luks +_require_drivers raw _make_test_img 64M diff --git a/tests/qemu-iotests/162 b/tests/qemu-iotests/162 index XXXXXXX..XXXXXXX 100755 --- a/tests/qemu-iotests/162 +++ b/tests/qemu-iotests/162 @@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15 . ./common.filter _supported_fmt generic - -test_ssh=$($QEMU_IMG --help | grep '^Supported formats:.* ssh\( \|$\)') -[ "$test_ssh" = "" ] && _notrun "ssh support required" +_require_drivers ssh echo echo '=== NBD ===' diff --git a/tests/qemu-iotests/184 b/tests/qemu-iotests/184 index XXXXXXX..XXXXXXX 100755 --- a/tests/qemu-iotests/184 +++ b/tests/qemu-iotests/184 @@ -XXX,XX +XXX,XX @@ trap "exit \$status" 0 1 2 3 15 . ./common.filter _supported_os Linux +_require_drivers throttle do_run_qemu() { diff --git a/tests/qemu-iotests/186 b/tests/qemu-iotests/186 index XXXXXXX..XXXXXXX 100755 --- a/tests/qemu-iotests/186 +++ b/tests/qemu-iotests/186 @@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15 _supported_fmt qcow2 _supported_proto file +_require_drivers null-co if [ "$QEMU_DEFAULT_MACHINE" != "pc" ]; then _notrun "Requires a PC machine" diff --git a/tests/qemu-iotests/common.rc b/tests/qemu-iotests/common.rc index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/common.rc +++ b/tests/qemu-iotests/common.rc @@ -XXX,XX +XXX,XX @@ _require_command() [ -x "$c" ] || _notrun "$1 utility required, skipped this test" } +# Check that a set of drivers has been whitelisted in the QEMU binary +# +_require_drivers() +{ + available=$($QEMU -drive format=help | \ + sed -e '/Supported formats:/!d' -e 's/Supported formats://') + for driver + do + if ! echo "$available" | grep -q " $driver\( \|$\)"; then + _notrun "$driver not available" + fi + done +} + # make sure this script returns success true -- 2.21.0 From: Thomas Huth <thuth@redhat.com> The sanitizers (especially the address sanitizer from Clang) are sometimes printing out warnings or false positives - this spoils the output of the iotests, causing some of the tests to fail. Thus let's skip the automatic iotests during "make check" when the user configured QEMU with --enable-sanitizers. Signed-off-by: Thomas Huth <thuth@redhat.com> Message-id: 20190823084203.29734-1-thuth@redhat.com Signed-off-by: Max Reitz <mreitz@redhat.com> --- tests/check-block.sh | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/tests/check-block.sh b/tests/check-block.sh index XXXXXXX..XXXXXXX 100755 --- a/tests/check-block.sh +++ b/tests/check-block.sh @@ -XXX,XX +XXX,XX @@ if grep -q "TARGET_GPROF=y" *-softmmu/config-target.mak 2>/dev/null ; then exit 0 fi +if grep -q "CFLAGS.*-fsanitize" config-host.mak 2>/dev/null ; then + echo "Sanitizers are enabled ==> Not running the qemu-iotests." + exit 0 +fi + if [ -z "$(find . -name 'qemu-system-*' -print)" ]; then echo "No qemu-system binary available ==> Not running the qemu-iotests." exit 0 -- 2.21.0 From: Nir Soffer <nirsof@gmail.com> Quoting cache mode is not needed, and most tests use unquoted values. Unify all test to use the same style. Message-id: 20190827173432.7656-1-nsoffer@redhat.com Signed-off-by: Nir Soffer <nsoffer@redhat.com> Signed-off-by: Max Reitz <mreitz@redhat.com> --- tests/qemu-iotests/026 | 4 ++-- tests/qemu-iotests/039 | 4 ++-- tests/qemu-iotests/052 | 2 +- tests/qemu-iotests/091 | 4 ++-- 4 files changed, 7 insertions(+), 7 deletions(-) diff --git a/tests/qemu-iotests/026 b/tests/qemu-iotests/026 index XXXXXXX..XXXXXXX 100755 --- a/tests/qemu-iotests/026 +++ b/tests/qemu-iotests/026 @@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15 # Currently only qcow2 supports rebasing _supported_fmt qcow2 _supported_proto file -_default_cache_mode "writethrough" -_supported_cache_modes "writethrough" "none" +_default_cache_mode writethrough +_supported_cache_modes writethrough none # The refcount table tests expect a certain minimum width for refcount entries # (so that the refcount table actually needs to grow); that minimum is 16 bits, # being the default refcount entry width. diff --git a/tests/qemu-iotests/039 b/tests/qemu-iotests/039 index XXXXXXX..XXXXXXX 100755 --- a/tests/qemu-iotests/039 +++ b/tests/qemu-iotests/039 @@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15 _supported_fmt qcow2 _supported_proto file _supported_os Linux -_default_cache_mode "writethrough" -_supported_cache_modes "writethrough" +_default_cache_mode writethrough +_supported_cache_modes writethrough size=128M diff --git a/tests/qemu-iotests/052 b/tests/qemu-iotests/052 index XXXXXXX..XXXXXXX 100755 --- a/tests/qemu-iotests/052 +++ b/tests/qemu-iotests/052 @@ -XXX,XX +XXX,XX @@ _supported_fmt generic _supported_proto file # Don't do O_DIRECT on tmpfs -_supported_cache_modes "writeback" "writethrough" "unsafe" +_supported_cache_modes writeback writethrough unsafe size=128M _make_test_img $size diff --git a/tests/qemu-iotests/091 b/tests/qemu-iotests/091 index XXXXXXX..XXXXXXX 100755 --- a/tests/qemu-iotests/091 +++ b/tests/qemu-iotests/091 @@ -XXX,XX +XXX,XX @@ trap "_cleanup; exit \$status" 0 1 2 3 15 _supported_fmt qcow2 _supported_proto file _supported_os Linux -_default_cache_mode "none" -_supported_cache_modes "writethrough" "none" "writeback" +_default_cache_mode none +_supported_cache_modes writethrough none writeback size=1G -- 2.21.0 The following changes since commit 9ac5df20f51fabcba0d902025df4bd7ea987c158: Merge remote-tracking branch 'remotes/pmaydell/tags/pull-target-arm-20200221-1' into staging (2020-02-21 16:18:38 +0000) are available in the Git repository at: https://github.com/stefanha/qemu.git tags/block-pull-request for you to fetch changes up to e5c59355ae9f724777c61c859292ec9db2c8c2ab: fuzz: add documentation to docs/devel/ (2020-02-22 08:26:48 +0000) ---------------------------------------------------------------- Pull request This pull request contains a virtio-blk/scsi performance optimization, event loop scalability improvements, and a qtest-based device fuzzing framework. I am including the fuzzing patches because I have reviewed them and Thomas Huth is currently away on leave. ---------------------------------------------------------------- Alexander Bulekov (22): softmmu: move vl.c to softmmu/ softmmu: split off vl.c:main() into main.c module: check module wasn't already initialized fuzz: add FUZZ_TARGET module type qtest: add qtest_server_send abstraction libqtest: add a layer of abstraction to send/recv libqtest: make bufwrite rely on the TransportOps qtest: add in-process incoming command handler libqos: rename i2c_send and i2c_recv libqos: split qos-test and libqos makefile vars libqos: move useful qos-test funcs to qos_external fuzz: add fuzzer skeleton exec: keep ram block across fork when using qtest main: keep rcu_atfork callback enabled for qtest fuzz: support for fork-based fuzzing. fuzz: add support for qos-assisted fuzz targets fuzz: add target/fuzz makefile rules fuzz: add configure flag --enable-fuzzing fuzz: add i440fx fuzz targets fuzz: add virtio-net fuzz target fuzz: add virtio-scsi fuzz target fuzz: add documentation to docs/devel/ Denis Plotnikov (1): virtio: increase virtqueue size for virtio-scsi and virtio-blk Paolo Bonzini (1): rcu_queue: add QSLIST functions Stefan Hajnoczi (7): aio-posix: avoid reacquiring rcu_read_lock() when polling util/async: make bh_aio_poll() O(1) aio-posix: fix use after leaving scope in aio_poll() aio-posix: don't pass ns timeout to epoll_wait() qemu/queue.h: add QLIST_SAFE_REMOVE() aio-posix: make AioHandler deletion O(1) aio-posix: make AioHandler dispatch O(1) with epoll MAINTAINERS | 11 +- Makefile | 15 +- Makefile.objs | 2 - Makefile.target | 19 ++- block.c | 5 +- chardev/spice.c | 4 +- configure | 39 +++++ docs/devel/fuzzing.txt | 116 ++++++++++++++ exec.c | 12 +- hw/block/virtio-blk.c | 2 +- hw/core/machine.c | 2 + hw/scsi/virtio-scsi.c | 2 +- include/block/aio.h | 26 ++- include/qemu/module.h | 4 +- include/qemu/queue.h | 32 +++- include/qemu/rcu_queue.h | 47 ++++++ include/sysemu/qtest.h | 4 + include/sysemu/sysemu.h | 4 + qtest.c | 31 +++- scripts/checkpatch.pl | 2 +- scripts/get_maintainer.pl | 3 +- softmmu/Makefile.objs | 3 + softmmu/main.c | 53 +++++++ vl.c => softmmu/vl.c | 48 +++--- tests/Makefile.include | 2 + tests/qtest/Makefile.include | 72 +++++---- tests/qtest/fuzz/Makefile.include | 18 +++ tests/qtest/fuzz/fork_fuzz.c | 55 +++++++ tests/qtest/fuzz/fork_fuzz.h | 23 +++ tests/qtest/fuzz/fork_fuzz.ld | 37 +++++ tests/qtest/fuzz/fuzz.c | 179 +++++++++++++++++++++ tests/qtest/fuzz/fuzz.h | 95 +++++++++++ tests/qtest/fuzz/i440fx_fuzz.c | 193 ++++++++++++++++++++++ tests/qtest/fuzz/qos_fuzz.c | 234 +++++++++++++++++++++++++++ tests/qtest/fuzz/qos_fuzz.h | 33 ++++ tests/qtest/fuzz/virtio_net_fuzz.c | 198 +++++++++++++++++++++++ tests/qtest/fuzz/virtio_scsi_fuzz.c | 213 +++++++++++++++++++++++++ tests/qtest/libqos/i2c.c | 10 +- tests/qtest/libqos/i2c.h | 4 +- tests/qtest/libqos/qos_external.c | 168 ++++++++++++++++++++ tests/qtest/libqos/qos_external.h | 28 ++++ tests/qtest/libqtest.c | 119 ++++++++++++-- tests/qtest/libqtest.h | 4 + tests/qtest/pca9552-test.c | 10 +- tests/qtest/qos-test.c | 132 +--------------- tests/test-aio.c | 3 +- tests/test-rcu-list.c | 16 ++ tests/test-rcu-slist.c | 2 + util/aio-posix.c | 187 +++++++++++++++------- util/async.c | 237 ++++++++++++++++------------ util/module.c | 7 + 51 files changed, 2365 insertions(+), 400 deletions(-) create mode 100644 docs/devel/fuzzing.txt create mode 100644 softmmu/Makefile.objs create mode 100644 softmmu/main.c rename vl.c => softmmu/vl.c (99%) create mode 100644 tests/qtest/fuzz/Makefile.include create mode 100644 tests/qtest/fuzz/fork_fuzz.c create mode 100644 tests/qtest/fuzz/fork_fuzz.h create mode 100644 tests/qtest/fuzz/fork_fuzz.ld create mode 100644 tests/qtest/fuzz/fuzz.c create mode 100644 tests/qtest/fuzz/fuzz.h create mode 100644 tests/qtest/fuzz/i440fx_fuzz.c create mode 100644 tests/qtest/fuzz/qos_fuzz.c create mode 100644 tests/qtest/fuzz/qos_fuzz.h create mode 100644 tests/qtest/fuzz/virtio_net_fuzz.c create mode 100644 tests/qtest/fuzz/virtio_scsi_fuzz.c create mode 100644 tests/qtest/libqos/qos_external.c create mode 100644 tests/qtest/libqos/qos_external.h create mode 100644 tests/test-rcu-slist.c -- 2.24.1 From: Denis Plotnikov <dplotnikov@virtuozzo.com> The goal is to reduce the amount of requests issued by a guest on 1M reads/writes. This rises the performance up to 4% on that kind of disk access pattern. The maximum chunk size to be used for the guest disk accessing is limited with seg_max parameter, which represents the max amount of pices in the scatter-geather list in one guest disk request. Since seg_max is virqueue_size dependent, increasing the virtqueue size increases seg_max, which, in turn, increases the maximum size of data to be read/write from a guest disk. More details in the original problem statment: https://lists.gnu.org/archive/html/qemu-devel/2017-12/msg03721.html Suggested-by: Denis V. Lunev <den@openvz.org> Signed-off-by: Denis Plotnikov <dplotnikov@virtuozzo.com> Message-id: 20200214074648.958-1-dplotnikov@virtuozzo.com Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> --- hw/block/virtio-blk.c | 2 +- hw/core/machine.c | 2 ++ hw/scsi/virtio-scsi.c | 2 +- 3 files changed, 4 insertions(+), 2 deletions(-) diff --git a/hw/block/virtio-blk.c b/hw/block/virtio-blk.c index XXXXXXX..XXXXXXX 100644 --- a/hw/block/virtio-blk.c +++ b/hw/block/virtio-blk.c @@ -XXX,XX +XXX,XX @@ static Property virtio_blk_properties[] = { DEFINE_PROP_BIT("request-merging", VirtIOBlock, conf.request_merging, 0, true), DEFINE_PROP_UINT16("num-queues", VirtIOBlock, conf.num_queues, 1), - DEFINE_PROP_UINT16("queue-size", VirtIOBlock, conf.queue_size, 128), + DEFINE_PROP_UINT16("queue-size", VirtIOBlock, conf.queue_size, 256), DEFINE_PROP_BOOL("seg-max-adjust", VirtIOBlock, conf.seg_max_adjust, true), DEFINE_PROP_LINK("iothread", VirtIOBlock, conf.iothread, TYPE_IOTHREAD, IOThread *), diff --git a/hw/core/machine.c b/hw/core/machine.c index XXXXXXX..XXXXXXX 100644 --- a/hw/core/machine.c +++ b/hw/core/machine.c @@ -XXX,XX +XXX,XX @@ #include "hw/mem/nvdimm.h" GlobalProperty hw_compat_4_2[] = { + { "virtio-blk-device", "queue-size", "128"}, + { "virtio-scsi-device", "virtqueue_size", "128"}, { "virtio-blk-device", "x-enable-wce-if-config-wce", "off" }, { "virtio-blk-device", "seg-max-adjust", "off"}, { "virtio-scsi-device", "seg_max_adjust", "off"}, diff --git a/hw/scsi/virtio-scsi.c b/hw/scsi/virtio-scsi.c index XXXXXXX..XXXXXXX 100644 --- a/hw/scsi/virtio-scsi.c +++ b/hw/scsi/virtio-scsi.c @@ -XXX,XX +XXX,XX @@ static void virtio_scsi_device_unrealize(DeviceState *dev, Error **errp) static Property virtio_scsi_properties[] = { DEFINE_PROP_UINT32("num_queues", VirtIOSCSI, parent_obj.conf.num_queues, 1), DEFINE_PROP_UINT32("virtqueue_size", VirtIOSCSI, - parent_obj.conf.virtqueue_size, 128), + parent_obj.conf.virtqueue_size, 256), DEFINE_PROP_BOOL("seg_max_adjust", VirtIOSCSI, parent_obj.conf.seg_max_adjust, true), DEFINE_PROP_UINT32("max_sectors", VirtIOSCSI, parent_obj.conf.max_sectors, -- 2.24.1 The first rcu_read_lock/unlock() is expensive. Nested calls are cheap. This optimization increases IOPS from 73k to 162k with a Linux guest that has 2 virtio-blk,num-queues=1 and 99 virtio-blk,num-queues=32 devices. Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Paolo Bonzini <pbonzini@redhat.com> Message-id: 20200218182708.914552-1-stefanha@redhat.com Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> --- util/aio-posix.c | 11 +++++++++++ 1 file changed, 11 insertions(+) diff --git a/util/aio-posix.c b/util/aio-posix.c index XXXXXXX..XXXXXXX 100644 --- a/util/aio-posix.c +++ b/util/aio-posix.c @@ -XXX,XX +XXX,XX @@ #include "qemu/osdep.h" #include "block/block.h" +#include "qemu/rcu.h" #include "qemu/rcu_queue.h" #include "qemu/sockets.h" #include "qemu/cutils.h" @@ -XXX,XX +XXX,XX @@ static bool run_poll_handlers_once(AioContext *ctx, int64_t *timeout) bool progress = false; AioHandler *node; + /* + * Optimization: ->io_poll() handlers often contain RCU read critical + * sections and we therefore see many rcu_read_lock() -> rcu_read_unlock() + * -> rcu_read_lock() -> ... sequences with expensive memory + * synchronization primitives. Make the entire polling loop an RCU + * critical section because nested rcu_read_lock()/rcu_read_unlock() calls + * are cheap. + */ + RCU_READ_LOCK_GUARD(); + QLIST_FOREACH_RCU(node, &ctx->aio_handlers, node) { if (!node->deleted && node->io_poll && aio_node_check(ctx, node->is_external) && -- 2.24.1 From: Paolo Bonzini <pbonzini@redhat.com> QSLIST is the only family of lists for which we do not have RCU-friendly accessors, add them. Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Message-id: 20200220103828.24525-1-pbonzini@redhat.com Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> --- include/qemu/queue.h | 15 +++++++++++-- include/qemu/rcu_queue.h | 47 ++++++++++++++++++++++++++++++++++++++++ tests/Makefile.include | 2 ++ tests/test-rcu-list.c | 16 ++++++++++++++ tests/test-rcu-slist.c | 2 ++ 5 files changed, 80 insertions(+), 2 deletions(-) create mode 100644 tests/test-rcu-slist.c diff --git a/include/qemu/queue.h b/include/qemu/queue.h index XXXXXXX..XXXXXXX 100644 --- a/include/qemu/queue.h +++ b/include/qemu/queue.h @@ -XXX,XX +XXX,XX @@ struct { \ (head)->slh_first = (head)->slh_first->field.sle_next; \ } while (/*CONSTCOND*/0) -#define QSLIST_REMOVE_AFTER(slistelm, field) do { \ +#define QSLIST_REMOVE_AFTER(slistelm, field) do { \ (slistelm)->field.sle_next = \ - QSLIST_NEXT(QSLIST_NEXT((slistelm), field), field); \ + QSLIST_NEXT(QSLIST_NEXT((slistelm), field), field); \ +} while (/*CONSTCOND*/0) + +#define QSLIST_REMOVE(head, elm, type, field) do { \ + if ((head)->slh_first == (elm)) { \ + QSLIST_REMOVE_HEAD((head), field); \ + } else { \ + struct type *curelm = (head)->slh_first; \ + while (curelm->field.sle_next != (elm)) \ + curelm = curelm->field.sle_next; \ + curelm->field.sle_next = curelm->field.sle_next->field.sle_next; \ + } \ } while (/*CONSTCOND*/0) #define QSLIST_FOREACH(var, head, field) \ diff --git a/include/qemu/rcu_queue.h b/include/qemu/rcu_queue.h index XXXXXXX..XXXXXXX 100644 --- a/include/qemu/rcu_queue.h +++ b/include/qemu/rcu_queue.h @@ -XXX,XX +XXX,XX @@ extern "C" { (var) && ((next) = atomic_rcu_read(&(var)->field.tqe_next), 1); \ (var) = (next)) +/* + * RCU singly-linked list + */ + +/* Singly-linked list access methods */ +#define QSLIST_EMPTY_RCU(head) (atomic_read(&(head)->slh_first) == NULL) +#define QSLIST_FIRST_RCU(head) atomic_rcu_read(&(head)->slh_first) +#define QSLIST_NEXT_RCU(elm, field) atomic_rcu_read(&(elm)->field.sle_next) + +/* Singly-linked list functions */ +#define QSLIST_INSERT_HEAD_RCU(head, elm, field) do { \ + (elm)->field.sle_next = (head)->slh_first; \ + atomic_rcu_set(&(head)->slh_first, (elm)); \ +} while (/*CONSTCOND*/0) + +#define QSLIST_INSERT_AFTER_RCU(head, listelm, elm, field) do { \ + (elm)->field.sle_next = (listelm)->field.sle_next; \ + atomic_rcu_set(&(listelm)->field.sle_next, (elm)); \ +} while (/*CONSTCOND*/0) + +#define QSLIST_REMOVE_HEAD_RCU(head, field) do { \ + atomic_set(&(head)->slh_first, (head)->slh_first->field.sle_next); \ +} while (/*CONSTCOND*/0) + +#define QSLIST_REMOVE_RCU(head, elm, type, field) do { \ + if ((head)->slh_first == (elm)) { \ + QSLIST_REMOVE_HEAD_RCU((head), field); \ + } else { \ + struct type *curr = (head)->slh_first; \ + while (curr->field.sle_next != (elm)) { \ + curr = curr->field.sle_next; \ + } \ + atomic_set(&curr->field.sle_next, \ + curr->field.sle_next->field.sle_next); \ + } \ +} while (/*CONSTCOND*/0) + +#define QSLIST_FOREACH_RCU(var, head, field) \ + for ((var) = atomic_rcu_read(&(head)->slh_first); \ + (var); \ + (var) = atomic_rcu_read(&(var)->field.sle_next)) + +#define QSLIST_FOREACH_SAFE_RCU(var, head, field, next) \ + for ((var) = atomic_rcu_read(&(head)->slh_first); \ + (var) && ((next) = atomic_rcu_read(&(var)->field.sle_next), 1); \ + (var) = (next)) + #ifdef __cplusplus } #endif diff --git a/tests/Makefile.include b/tests/Makefile.include index XXXXXXX..XXXXXXX 100644 --- a/tests/Makefile.include +++ b/tests/Makefile.include @@ -XXX,XX +XXX,XX @@ check-unit-y += tests/rcutorture$(EXESUF) check-unit-y += tests/test-rcu-list$(EXESUF) check-unit-y += tests/test-rcu-simpleq$(EXESUF) check-unit-y += tests/test-rcu-tailq$(EXESUF) +check-unit-y += tests/test-rcu-slist$(EXESUF) check-unit-y += tests/test-qdist$(EXESUF) check-unit-y += tests/test-qht$(EXESUF) check-unit-y += tests/test-qht-par$(EXESUF) @@ -XXX,XX +XXX,XX @@ tests/rcutorture$(EXESUF): tests/rcutorture.o $(test-util-obj-y) tests/test-rcu-list$(EXESUF): tests/test-rcu-list.o $(test-util-obj-y) tests/test-rcu-simpleq$(EXESUF): tests/test-rcu-simpleq.o $(test-util-obj-y) tests/test-rcu-tailq$(EXESUF): tests/test-rcu-tailq.o $(test-util-obj-y) +tests/test-rcu-slist$(EXESUF): tests/test-rcu-slist.o $(test-util-obj-y) tests/test-qdist$(EXESUF): tests/test-qdist.o $(test-util-obj-y) tests/test-qht$(EXESUF): tests/test-qht.o $(test-util-obj-y) tests/test-qht-par$(EXESUF): tests/test-qht-par.o tests/qht-bench$(EXESUF) $(test-util-obj-y) diff --git a/tests/test-rcu-list.c b/tests/test-rcu-list.c index XXXXXXX..XXXXXXX 100644 --- a/tests/test-rcu-list.c +++ b/tests/test-rcu-list.c @@ -XXX,XX +XXX,XX @@ struct list_element { QSIMPLEQ_ENTRY(list_element) entry; #elif TEST_LIST_TYPE == 3 QTAILQ_ENTRY(list_element) entry; +#elif TEST_LIST_TYPE == 4 + QSLIST_ENTRY(list_element) entry; #else #error Invalid TEST_LIST_TYPE #endif @@ -XXX,XX +XXX,XX @@ static QTAILQ_HEAD(, list_element) Q_list_head; #define TEST_LIST_INSERT_HEAD_RCU QTAILQ_INSERT_HEAD_RCU #define TEST_LIST_FOREACH_RCU QTAILQ_FOREACH_RCU #define TEST_LIST_FOREACH_SAFE_RCU QTAILQ_FOREACH_SAFE_RCU + +#elif TEST_LIST_TYPE == 4 +static QSLIST_HEAD(, list_element) Q_list_head; + +#define TEST_NAME "qslist" +#define TEST_LIST_REMOVE_RCU(el, f) \ + QSLIST_REMOVE_RCU(&Q_list_head, el, list_element, f) + +#define TEST_LIST_INSERT_AFTER_RCU(list_el, el, f) \ + QSLIST_INSERT_AFTER_RCU(&Q_list_head, list_el, el, f) + +#define TEST_LIST_INSERT_HEAD_RCU QSLIST_INSERT_HEAD_RCU +#define TEST_LIST_FOREACH_RCU QSLIST_FOREACH_RCU +#define TEST_LIST_FOREACH_SAFE_RCU QSLIST_FOREACH_SAFE_RCU #else #error Invalid TEST_LIST_TYPE #endif diff --git a/tests/test-rcu-slist.c b/tests/test-rcu-slist.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/tests/test-rcu-slist.c @@ -XXX,XX +XXX,XX @@ +#define TEST_LIST_TYPE 4 +#include "test-rcu-list.c" -- 2.24.1 The ctx->first_bh list contains all created BHs, including those that are not scheduled. The list is iterated by the event loop and therefore has O(n) time complexity with respected to the number of created BHs. Rewrite BHs so that only scheduled or deleted BHs are enqueued. Only BHs that actually require action will be iterated. One semantic change is required: qemu_bh_delete() enqueues the BH and therefore invokes aio_notify(). The tests/test-aio.c:test_source_bh_delete_from_cb() test case assumed that g_main_context_iteration(NULL, false) returns false after qemu_bh_delete() but it now returns true for one iteration. Fix up the test case. This patch makes aio_compute_timeout() and aio_bh_poll() drop from a CPU profile reported by perf-top(1). Previously they combined to 9% CPU utilization when AioContext polling is commented out and the guest has 2 virtio-blk,num-queues=1 and 99 virtio-blk,num-queues=32 devices. Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Paolo Bonzini <pbonzini@redhat.com> Message-id: 20200221093951.1414693-1-stefanha@redhat.com Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> --- include/block/aio.h | 20 +++- tests/test-aio.c | 3 +- util/async.c | 237 ++++++++++++++++++++++++++------------------ 3 files changed, 158 insertions(+), 102 deletions(-) diff --git a/include/block/aio.h b/include/block/aio.h index XXXXXXX..XXXXXXX 100644 --- a/include/block/aio.h +++ b/include/block/aio.h @@ -XXX,XX +XXX,XX @@ struct ThreadPool; struct LinuxAioState; struct LuringState; +/* + * Each aio_bh_poll() call carves off a slice of the BH list, so that newly + * scheduled BHs are not processed until the next aio_bh_poll() call. All + * active aio_bh_poll() calls chain their slices together in a list, so that + * nested aio_bh_poll() calls process all scheduled bottom halves. + */ +typedef QSLIST_HEAD(, QEMUBH) BHList; +typedef struct BHListSlice BHListSlice; +struct BHListSlice { + BHList bh_list; + QSIMPLEQ_ENTRY(BHListSlice) next; +}; + struct AioContext { GSource source; @@ -XXX,XX +XXX,XX @@ struct AioContext { */ QemuLockCnt list_lock; - /* Anchor of the list of Bottom Halves belonging to the context */ - struct QEMUBH *first_bh; + /* Bottom Halves pending aio_bh_poll() processing */ + BHList bh_list; + + /* Chained BH list slices for each nested aio_bh_poll() call */ + QSIMPLEQ_HEAD(, BHListSlice) bh_slice_list; /* Used by aio_notify. * diff --git a/tests/test-aio.c b/tests/test-aio.c index XXXXXXX..XXXXXXX 100644 --- a/tests/test-aio.c +++ b/tests/test-aio.c @@ -XXX,XX +XXX,XX @@ static void test_source_bh_delete_from_cb(void) g_assert_cmpint(data1.n, ==, data1.max); g_assert(data1.bh == NULL); - g_assert(!g_main_context_iteration(NULL, false)); + assert(g_main_context_iteration(NULL, false)); + assert(!g_main_context_iteration(NULL, false)); } static void test_source_bh_delete_from_cb_many(void) diff --git a/util/async.c b/util/async.c index XXXXXXX..XXXXXXX 100644 --- a/util/async.c +++ b/util/async.c @@ -XXX,XX +XXX,XX @@ #include "block/thread-pool.h" #include "qemu/main-loop.h" #include "qemu/atomic.h" +#include "qemu/rcu_queue.h" #include "block/raw-aio.h" #include "qemu/coroutine_int.h" #include "trace.h" @@ -XXX,XX +XXX,XX @@ /***********************************************************/ /* bottom halves (can be seen as timers which expire ASAP) */ +/* QEMUBH::flags values */ +enum { + /* Already enqueued and waiting for aio_bh_poll() */ + BH_PENDING = (1 << 0), + + /* Invoke the callback */ + BH_SCHEDULED = (1 << 1), + + /* Delete without invoking callback */ + BH_DELETED = (1 << 2), + + /* Delete after invoking callback */ + BH_ONESHOT = (1 << 3), + + /* Schedule periodically when the event loop is idle */ + BH_IDLE = (1 << 4), +}; + struct QEMUBH { AioContext *ctx; QEMUBHFunc *cb; void *opaque; - QEMUBH *next; - bool scheduled; - bool idle; - bool deleted; + QSLIST_ENTRY(QEMUBH) next; + unsigned flags; }; +/* Called concurrently from any thread */ +static void aio_bh_enqueue(QEMUBH *bh, unsigned new_flags) +{ + AioContext *ctx = bh->ctx; + unsigned old_flags; + + /* + * The memory barrier implicit in atomic_fetch_or makes sure that: + * 1. idle & any writes needed by the callback are done before the + * locations are read in the aio_bh_poll. + * 2. ctx is loaded before the callback has a chance to execute and bh + * could be freed. + */ + old_flags = atomic_fetch_or(&bh->flags, BH_PENDING | new_flags); + if (!(old_flags & BH_PENDING)) { + QSLIST_INSERT_HEAD_ATOMIC(&ctx->bh_list, bh, next); + } + + aio_notify(ctx); +} + +/* Only called from aio_bh_poll() and aio_ctx_finalize() */ +static QEMUBH *aio_bh_dequeue(BHList *head, unsigned *flags) +{ + QEMUBH *bh = QSLIST_FIRST_RCU(head); + + if (!bh) { + return NULL; + } + + QSLIST_REMOVE_HEAD(head, next); + + /* + * The atomic_and is paired with aio_bh_enqueue(). The implicit memory + * barrier ensures that the callback sees all writes done by the scheduling + * thread. It also ensures that the scheduling thread sees the cleared + * flag before bh->cb has run, and thus will call aio_notify again if + * necessary. + */ + *flags = atomic_fetch_and(&bh->flags, + ~(BH_PENDING | BH_SCHEDULED | BH_IDLE)); + return bh; +} + void aio_bh_schedule_oneshot(AioContext *ctx, QEMUBHFunc *cb, void *opaque) { QEMUBH *bh; @@ -XXX,XX +XXX,XX @@ void aio_bh_schedule_oneshot(AioContext *ctx, QEMUBHFunc *cb, void *opaque) .cb = cb, .opaque = opaque, }; - qemu_lockcnt_lock(&ctx->list_lock); - bh->next = ctx->first_bh; - bh->scheduled = 1; - bh->deleted = 1; - /* Make sure that the members are ready before putting bh into list */ - smp_wmb(); - ctx->first_bh = bh; - qemu_lockcnt_unlock(&ctx->list_lock); - aio_notify(ctx); + aio_bh_enqueue(bh, BH_SCHEDULED | BH_ONESHOT); } QEMUBH *aio_bh_new(AioContext *ctx, QEMUBHFunc *cb, void *opaque) @@ -XXX,XX +XXX,XX @@ QEMUBH *aio_bh_new(AioContext *ctx, QEMUBHFunc *cb, void *opaque) .cb = cb, .opaque = opaque, }; - qemu_lockcnt_lock(&ctx->list_lock); - bh->next = ctx->first_bh; - /* Make sure that the members are ready before putting bh into list */ - smp_wmb(); - ctx->first_bh = bh; - qemu_lockcnt_unlock(&ctx->list_lock); return bh; } @@ -XXX,XX +XXX,XX @@ void aio_bh_call(QEMUBH *bh) bh->cb(bh->opaque); } -/* Multiple occurrences of aio_bh_poll cannot be called concurrently. - * The count in ctx->list_lock is incremented before the call, and is - * not affected by the call. - */ +/* Multiple occurrences of aio_bh_poll cannot be called concurrently. */ int aio_bh_poll(AioContext *ctx) { - QEMUBH *bh, **bhp, *next; - int ret; - bool deleted = false; - - ret = 0; - for (bh = atomic_rcu_read(&ctx->first_bh); bh; bh = next) { - next = atomic_rcu_read(&bh->next); - /* The atomic_xchg is paired with the one in qemu_bh_schedule. The - * implicit memory barrier ensures that the callback sees all writes - * done by the scheduling thread. It also ensures that the scheduling - * thread sees the zero before bh->cb has run, and thus will call - * aio_notify again if necessary. - */ - if (atomic_xchg(&bh->scheduled, 0)) { + BHListSlice slice; + BHListSlice *s; + int ret = 0; + + QSLIST_MOVE_ATOMIC(&slice.bh_list, &ctx->bh_list); + QSIMPLEQ_INSERT_TAIL(&ctx->bh_slice_list, &slice, next); + + while ((s = QSIMPLEQ_FIRST(&ctx->bh_slice_list))) { + QEMUBH *bh; + unsigned flags; + + bh = aio_bh_dequeue(&s->bh_list, &flags); + if (!bh) { + QSIMPLEQ_REMOVE_HEAD(&ctx->bh_slice_list, next); + continue; + } + + if ((flags & (BH_SCHEDULED | BH_DELETED)) == BH_SCHEDULED) { /* Idle BHs don't count as progress */ - if (!bh->idle) { + if (!(flags & BH_IDLE)) { ret = 1; } - bh->idle = 0; aio_bh_call(bh); } - if (bh->deleted) { - deleted = true; + if (flags & (BH_DELETED | BH_ONESHOT)) { + g_free(bh); } } - /* remove deleted bhs */ - if (!deleted) { - return ret; - } - - if (qemu_lockcnt_dec_if_lock(&ctx->list_lock)) { - bhp = &ctx->first_bh; - while (*bhp) { - bh = *bhp; - if (bh->deleted && !bh->scheduled) { - *bhp = bh->next; - g_free(bh); - } else { - bhp = &bh->next; - } - } - qemu_lockcnt_inc_and_unlock(&ctx->list_lock); - } return ret; } void qemu_bh_schedule_idle(QEMUBH *bh) { - bh->idle = 1; - /* Make sure that idle & any writes needed by the callback are done - * before the locations are read in the aio_bh_poll. - */ - atomic_mb_set(&bh->scheduled, 1); + aio_bh_enqueue(bh, BH_SCHEDULED | BH_IDLE); } void qemu_bh_schedule(QEMUBH *bh) { - AioContext *ctx; - - ctx = bh->ctx; - bh->idle = 0; - /* The memory barrier implicit in atomic_xchg makes sure that: - * 1. idle & any writes needed by the callback are done before the - * locations are read in the aio_bh_poll. - * 2. ctx is loaded before scheduled is set and the callback has a chance - * to execute. - */ - if (atomic_xchg(&bh->scheduled, 1) == 0) { - aio_notify(ctx); - } + aio_bh_enqueue(bh, BH_SCHEDULED); } - /* This func is async. */ void qemu_bh_cancel(QEMUBH *bh) { - atomic_mb_set(&bh->scheduled, 0); + atomic_and(&bh->flags, ~BH_SCHEDULED); } /* This func is async.The bottom half will do the delete action at the finial @@ -XXX,XX +XXX,XX @@ void qemu_bh_cancel(QEMUBH *bh) */ void qemu_bh_delete(QEMUBH *bh) { - bh->scheduled = 0; - bh->deleted = 1; + aio_bh_enqueue(bh, BH_DELETED); } -int64_t -aio_compute_timeout(AioContext *ctx) +static int64_t aio_compute_bh_timeout(BHList *head, int timeout) { - int64_t deadline; - int timeout = -1; QEMUBH *bh; - for (bh = atomic_rcu_read(&ctx->first_bh); bh; - bh = atomic_rcu_read(&bh->next)) { - if (bh->scheduled) { - if (bh->idle) { + QSLIST_FOREACH_RCU(bh, head, next) { + if ((bh->flags & (BH_SCHEDULED | BH_DELETED)) == BH_SCHEDULED) { + if (bh->flags & BH_IDLE) { /* idle bottom halves will be polled at least * every 10ms */ timeout = 10000000; @@ -XXX,XX +XXX,XX @@ aio_compute_timeout(AioContext *ctx) } } + return timeout; +} + +int64_t +aio_compute_timeout(AioContext *ctx) +{ + BHListSlice *s; + int64_t deadline; + int timeout = -1; + + timeout = aio_compute_bh_timeout(&ctx->bh_list, timeout); + if (timeout == 0) { + return 0; + } + + QSIMPLEQ_FOREACH(s, &ctx->bh_slice_list, next) { + timeout = aio_compute_bh_timeout(&s->bh_list, timeout); + if (timeout == 0) { + return 0; + } + } + deadline = timerlistgroup_deadline_ns(&ctx->tlg); if (deadline == 0) { return 0; @@ -XXX,XX +XXX,XX @@ aio_ctx_check(GSource *source) { AioContext *ctx = (AioContext *) source; QEMUBH *bh; + BHListSlice *s; atomic_and(&ctx->notify_me, ~1); aio_notify_accept(ctx); - for (bh = ctx->first_bh; bh; bh = bh->next) { - if (bh->scheduled) { + QSLIST_FOREACH_RCU(bh, &ctx->bh_list, next) { + if ((bh->flags & (BH_SCHEDULED | BH_DELETED)) == BH_SCHEDULED) { return true; } } + + QSIMPLEQ_FOREACH(s, &ctx->bh_slice_list, next) { + QSLIST_FOREACH_RCU(bh, &s->bh_list, next) { + if ((bh->flags & (BH_SCHEDULED | BH_DELETED)) == BH_SCHEDULED) { + return true; + } + } + } return aio_pending(ctx) || (timerlistgroup_deadline_ns(&ctx->tlg) == 0); } @@ -XXX,XX +XXX,XX @@ static void aio_ctx_finalize(GSource *source) { AioContext *ctx = (AioContext *) source; + QEMUBH *bh; + unsigned flags; thread_pool_free(ctx->thread_pool); @@ -XXX,XX +XXX,XX @@ aio_ctx_finalize(GSource *source) assert(QSLIST_EMPTY(&ctx->scheduled_coroutines)); qemu_bh_delete(ctx->co_schedule_bh); - qemu_lockcnt_lock(&ctx->list_lock); - assert(!qemu_lockcnt_count(&ctx->list_lock)); - while (ctx->first_bh) { - QEMUBH *next = ctx->first_bh->next; + /* There must be no aio_bh_poll() calls going on */ + assert(QSIMPLEQ_EMPTY(&ctx->bh_slice_list)); + while ((bh = aio_bh_dequeue(&ctx->bh_list, &flags))) { /* qemu_bh_delete() must have been called on BHs in this AioContext */ - assert(ctx->first_bh->deleted); + assert(flags & BH_DELETED); - g_free(ctx->first_bh); - ctx->first_bh = next; + g_free(bh); } - qemu_lockcnt_unlock(&ctx->list_lock); aio_set_event_notifier(ctx, &ctx->notifier, false, NULL, NULL); event_notifier_cleanup(&ctx->notifier); @@ -XXX,XX +XXX,XX @@ AioContext *aio_context_new(Error **errp) AioContext *ctx; ctx = (AioContext *) g_source_new(&aio_source_funcs, sizeof(AioContext)); + QSLIST_INIT(&ctx->bh_list); + QSIMPLEQ_INIT(&ctx->bh_slice_list); aio_context_setup(ctx); ret = event_notifier_init(&ctx->notifier, false); -- 2.24.1 epoll_handler is a stack variable and must not be accessed after it goes out of scope: if (aio_epoll_check_poll(ctx, pollfds, npfd, timeout)) { AioHandler epoll_handler; ... add_pollfd(&epoll_handler); ret = aio_epoll(ctx, pollfds, npfd, timeout); } ... ... /* if we have any readable fds, dispatch event */ if (ret > 0) { for (i = 0; i < npfd; i++) { nodes[i]->pfd.revents = pollfds[i].revents; } } nodes[0] is &epoll_handler, which has already gone out of scope. There is no need to use pollfds[] for epoll. We don't need an AioHandler for the epoll fd. Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Sergio Lopez <slp@redhat.com> Message-id: 20200214171712.541358-2-stefanha@redhat.com Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> --- util/aio-posix.c | 20 ++++++++------------ 1 file changed, 8 insertions(+), 12 deletions(-) diff --git a/util/aio-posix.c b/util/aio-posix.c index XXXXXXX..XXXXXXX 100644 --- a/util/aio-posix.c +++ b/util/aio-posix.c @@ -XXX,XX +XXX,XX @@ static void aio_epoll_update(AioContext *ctx, AioHandler *node, bool is_new) } } -static int aio_epoll(AioContext *ctx, GPollFD *pfds, - unsigned npfd, int64_t timeout) +static int aio_epoll(AioContext *ctx, int64_t timeout) { + GPollFD pfd = { + .fd = ctx->epollfd, + .events = G_IO_IN | G_IO_OUT | G_IO_HUP | G_IO_ERR, + }; AioHandler *node; int i, ret = 0; struct epoll_event events[128]; - assert(npfd == 1); - assert(pfds[0].fd == ctx->epollfd); if (timeout > 0) { - ret = qemu_poll_ns(pfds, npfd, timeout); + ret = qemu_poll_ns(&pfd, 1, timeout); } if (timeout <= 0 || ret > 0) { ret = epoll_wait(ctx->epollfd, events, @@ -XXX,XX +XXX,XX @@ bool aio_poll(AioContext *ctx, bool blocking) /* wait until next event */ if (aio_epoll_check_poll(ctx, pollfds, npfd, timeout)) { - AioHandler epoll_handler; - - epoll_handler.pfd.fd = ctx->epollfd; - epoll_handler.pfd.events = G_IO_IN | G_IO_OUT | G_IO_HUP | G_IO_ERR; - npfd = 0; - add_pollfd(&epoll_handler); - ret = aio_epoll(ctx, pollfds, npfd, timeout); + npfd = 0; /* pollfds[] is not being used */ + ret = aio_epoll(ctx, timeout); } else { ret = qemu_poll_ns(pollfds, npfd, timeout); } -- 2.24.1 Don't pass the nanosecond timeout into epoll_wait(), which expects milliseconds. The epoll_wait() timeout value does not matter if qemu_poll_ns() determined that the poll fd is ready, but passing a value in the wrong units is still ugly. Pass a 0 timeout to epoll_wait() instead. Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Sergio Lopez <slp@redhat.com> Message-id: 20200214171712.541358-3-stefanha@redhat.com Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> --- util/aio-posix.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/util/aio-posix.c b/util/aio-posix.c index XXXXXXX..XXXXXXX 100644 --- a/util/aio-posix.c +++ b/util/aio-posix.c @@ -XXX,XX +XXX,XX @@ static int aio_epoll(AioContext *ctx, int64_t timeout) if (timeout > 0) { ret = qemu_poll_ns(&pfd, 1, timeout); + if (ret > 0) { + timeout = 0; + } } if (timeout <= 0 || ret > 0) { ret = epoll_wait(ctx->epollfd, events, -- 2.24.1 QLIST_REMOVE() assumes the element is in a list. It also leaves the element's linked list pointers dangling. Introduce a safe version of QLIST_REMOVE() and convert open-coded instances of this pattern. Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Sergio Lopez <slp@redhat.com> Message-id: 20200214171712.541358-4-stefanha@redhat.com Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> --- block.c | 5 +---- chardev/spice.c | 4 +--- include/qemu/queue.h | 14 ++++++++++++++ 3 files changed, 16 insertions(+), 7 deletions(-) diff --git a/block.c b/block.c index XXXXXXX..XXXXXXX 100644 --- a/block.c +++ b/block.c @@ -XXX,XX +XXX,XX @@ BdrvChild *bdrv_attach_child(BlockDriverState *parent_bs, static void bdrv_detach_child(BdrvChild *child) { - if (child->next.le_prev) { - QLIST_REMOVE(child, next); - child->next.le_prev = NULL; - } + QLIST_SAFE_REMOVE(child, next); bdrv_replace_child(child, NULL); diff --git a/chardev/spice.c b/chardev/spice.c index XXXXXXX..XXXXXXX 100644 --- a/chardev/spice.c +++ b/chardev/spice.c @@ -XXX,XX +XXX,XX @@ static void char_spice_finalize(Object *obj) vmc_unregister_interface(s); - if (s->next.le_prev) { - QLIST_REMOVE(s, next); - } + QLIST_SAFE_REMOVE(s, next); g_free((char *)s->sin.subtype); g_free((char *)s->sin.portname); diff --git a/include/qemu/queue.h b/include/qemu/queue.h index XXXXXXX..XXXXXXX 100644 --- a/include/qemu/queue.h +++ b/include/qemu/queue.h @@ -XXX,XX +XXX,XX @@ struct { \ *(elm)->field.le_prev = (elm)->field.le_next; \ } while (/*CONSTCOND*/0) +/* + * Like QLIST_REMOVE() but safe to call when elm is not in a list + */ +#define QLIST_SAFE_REMOVE(elm, field) do { \ + if ((elm)->field.le_prev != NULL) { \ + if ((elm)->field.le_next != NULL) \ + (elm)->field.le_next->field.le_prev = \ + (elm)->field.le_prev; \ + *(elm)->field.le_prev = (elm)->field.le_next; \ + (elm)->field.le_next = NULL; \ + (elm)->field.le_prev = NULL; \ + } \ +} while (/*CONSTCOND*/0) + #define QLIST_FOREACH(var, head, field) \ for ((var) = ((head)->lh_first); \ (var); \ -- 2.24.1 It is not necessary to scan all AioHandlers for deletion. Keep a list of deleted handlers instead of scanning the full list of all handlers. The AioHandler->deleted field can be dropped. Let's check if the handler has been inserted into the deleted list instead. Add a new QLIST_IS_INSERTED() API for this check. Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Sergio Lopez <slp@redhat.com> Message-id: 20200214171712.541358-5-stefanha@redhat.com Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> --- include/block/aio.h | 6 ++++- include/qemu/queue.h | 3 +++ util/aio-posix.c | 53 +++++++++++++++++++++++++++++--------------- 3 files changed, 43 insertions(+), 19 deletions(-) diff --git a/include/block/aio.h b/include/block/aio.h index XXXXXXX..XXXXXXX 100644 --- a/include/block/aio.h +++ b/include/block/aio.h @@ -XXX,XX +XXX,XX @@ void qemu_aio_unref(void *p); void qemu_aio_ref(void *p); typedef struct AioHandler AioHandler; +typedef QLIST_HEAD(, AioHandler) AioHandlerList; typedef void QEMUBHFunc(void *opaque); typedef bool AioPollFn(void *opaque); typedef void IOHandler(void *opaque); @@ -XXX,XX +XXX,XX @@ struct AioContext { QemuRecMutex lock; /* The list of registered AIO handlers. Protected by ctx->list_lock. */ - QLIST_HEAD(, AioHandler) aio_handlers; + AioHandlerList aio_handlers; + + /* The list of AIO handlers to be deleted. Protected by ctx->list_lock. */ + AioHandlerList deleted_aio_handlers; /* Used to avoid unnecessary event_notifier_set calls in aio_notify; * accessed with atomic primitives. If this field is 0, everything diff --git a/include/qemu/queue.h b/include/qemu/queue.h index XXXXXXX..XXXXXXX 100644 --- a/include/qemu/queue.h +++ b/include/qemu/queue.h @@ -XXX,XX +XXX,XX @@ struct { \ } \ } while (/*CONSTCOND*/0) +/* Is elm in a list? */ +#define QLIST_IS_INSERTED(elm, field) ((elm)->field.le_prev != NULL) + #define QLIST_FOREACH(var, head, field) \ for ((var) = ((head)->lh_first); \ (var); \ diff --git a/util/aio-posix.c b/util/aio-posix.c index XXXXXXX..XXXXXXX 100644 --- a/util/aio-posix.c +++ b/util/aio-posix.c @@ -XXX,XX +XXX,XX @@ struct AioHandler AioPollFn *io_poll; IOHandler *io_poll_begin; IOHandler *io_poll_end; - int deleted; void *opaque; bool is_external; QLIST_ENTRY(AioHandler) node; + QLIST_ENTRY(AioHandler) node_deleted; }; #ifdef CONFIG_EPOLL_CREATE1 @@ -XXX,XX +XXX,XX @@ static bool aio_epoll_try_enable(AioContext *ctx) QLIST_FOREACH_RCU(node, &ctx->aio_handlers, node) { int r; - if (node->deleted || !node->pfd.events) { + if (QLIST_IS_INSERTED(node, node_deleted) || !node->pfd.events) { continue; } event.events = epoll_events_from_pfd(node->pfd.events); @@ -XXX,XX +XXX,XX @@ static AioHandler *find_aio_handler(AioContext *ctx, int fd) AioHandler *node; QLIST_FOREACH(node, &ctx->aio_handlers, node) { - if (node->pfd.fd == fd) - if (!node->deleted) + if (node->pfd.fd == fd) { + if (!QLIST_IS_INSERTED(node, node_deleted)) { return node; + } + } } return NULL; @@ -XXX,XX +XXX,XX @@ static bool aio_remove_fd_handler(AioContext *ctx, AioHandler *node) /* If a read is in progress, just mark the node as deleted */ if (qemu_lockcnt_count(&ctx->list_lock)) { - node->deleted = 1; + QLIST_INSERT_HEAD_RCU(&ctx->deleted_aio_handlers, node, node_deleted); node->pfd.revents = 0; return false; } @@ -XXX,XX +XXX,XX @@ static void poll_set_started(AioContext *ctx, bool started) QLIST_FOREACH_RCU(node, &ctx->aio_handlers, node) { IOHandler *fn; - if (node->deleted) { + if (QLIST_IS_INSERTED(node, node_deleted)) { continue; } @@ -XXX,XX +XXX,XX @@ bool aio_pending(AioContext *ctx) return result; } +static void aio_free_deleted_handlers(AioContext *ctx) +{ + AioHandler *node; + + if (QLIST_EMPTY_RCU(&ctx->deleted_aio_handlers)) { + return; + } + if (!qemu_lockcnt_dec_if_lock(&ctx->list_lock)) { + return; /* we are nested, let the parent do the freeing */ + } + + while ((node = QLIST_FIRST_RCU(&ctx->deleted_aio_handlers))) { + QLIST_REMOVE(node, node); + QLIST_REMOVE(node, node_deleted); + g_free(node); + } + + qemu_lockcnt_inc_and_unlock(&ctx->list_lock); +} + static bool aio_dispatch_handlers(AioContext *ctx) { AioHandler *node, *tmp; @@ -XXX,XX +XXX,XX @@ static bool aio_dispatch_handlers(AioContext *ctx) revents = node->pfd.revents & node->pfd.events; node->pfd.revents = 0; - if (!node->deleted && + if (!QLIST_IS_INSERTED(node, node_deleted) && (revents & (G_IO_IN | G_IO_HUP | G_IO_ERR)) && aio_node_check(ctx, node->is_external) && node->io_read) { @@ -XXX,XX +XXX,XX @@ static bool aio_dispatch_handlers(AioContext *ctx) progress = true; } } - if (!node->deleted && + if (!QLIST_IS_INSERTED(node, node_deleted) && (revents & (G_IO_OUT | G_IO_ERR)) && aio_node_check(ctx, node->is_external) && node->io_write) { node->io_write(node->opaque); progress = true; } - - if (node->deleted) { - if (qemu_lockcnt_dec_if_lock(&ctx->list_lock)) { - QLIST_REMOVE(node, node); - g_free(node); - qemu_lockcnt_inc_and_unlock(&ctx->list_lock); - } - } } return progress; @@ -XXX,XX +XXX,XX @@ void aio_dispatch(AioContext *ctx) qemu_lockcnt_inc(&ctx->list_lock); aio_bh_poll(ctx); aio_dispatch_handlers(ctx); + aio_free_deleted_handlers(ctx); qemu_lockcnt_dec(&ctx->list_lock); timerlistgroup_run_timers(&ctx->tlg); @@ -XXX,XX +XXX,XX @@ static bool run_poll_handlers_once(AioContext *ctx, int64_t *timeout) RCU_READ_LOCK_GUARD(); QLIST_FOREACH_RCU(node, &ctx->aio_handlers, node) { - if (!node->deleted && node->io_poll && + if (!QLIST_IS_INSERTED(node, node_deleted) && node->io_poll && aio_node_check(ctx, node->is_external) && node->io_poll(node->opaque)) { /* @@ -XXX,XX +XXX,XX @@ bool aio_poll(AioContext *ctx, bool blocking) if (!aio_epoll_enabled(ctx)) { QLIST_FOREACH_RCU(node, &ctx->aio_handlers, node) { - if (!node->deleted && node->pfd.events + if (!QLIST_IS_INSERTED(node, node_deleted) && node->pfd.events && aio_node_check(ctx, node->is_external)) { add_pollfd(node); } @@ -XXX,XX +XXX,XX @@ bool aio_poll(AioContext *ctx, bool blocking) progress |= aio_dispatch_handlers(ctx); } + aio_free_deleted_handlers(ctx); + qemu_lockcnt_dec(&ctx->list_lock); progress |= timerlistgroup_run_timers(&ctx->tlg); -- 2.24.1 File descriptor monitoring is O(1) with epoll(7), but aio_dispatch_handlers() still scans all AioHandlers instead of dispatching just those that are ready. This makes aio_poll() O(n) with respect to the total number of registered handlers. Add a local ready_list to aio_poll() so that each nested aio_poll() builds a list of handlers ready to be dispatched. Since file descriptor polling is level-triggered, nested aio_poll() calls also see fds that were ready in the parent but not yet dispatched. This guarantees that nested aio_poll() invocations will dispatch all fds, even those that became ready before the nested invocation. Since only handlers ready to be dispatched are placed onto the ready_list, the new aio_dispatch_ready_handlers() function provides O(1) dispatch. Note that AioContext polling is still O(n) and currently cannot be fully disabled. This still needs to be fixed before aio_poll() is fully O(1). Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Sergio Lopez <slp@redhat.com> Message-id: 20200214171712.541358-6-stefanha@redhat.com [Fix compilation error on macOS where there is no epoll(87). The aio_epoll() prototype was out of date and aio_add_ready_list() needed to be moved outside the ifdef. --Stefan] Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> --- util/aio-posix.c | 110 +++++++++++++++++++++++++++++++++-------------- 1 file changed, 78 insertions(+), 32 deletions(-) diff --git a/util/aio-posix.c b/util/aio-posix.c index XXXXXXX..XXXXXXX 100644 --- a/util/aio-posix.c +++ b/util/aio-posix.c @@ -XXX,XX +XXX,XX @@ struct AioHandler void *opaque; bool is_external; QLIST_ENTRY(AioHandler) node; + QLIST_ENTRY(AioHandler) node_ready; /* only used during aio_poll() */ QLIST_ENTRY(AioHandler) node_deleted; }; +/* Add a handler to a ready list */ +static void add_ready_handler(AioHandlerList *ready_list, + AioHandler *node, + int revents) +{ + QLIST_SAFE_REMOVE(node, node_ready); /* remove from nested parent's list */ + node->pfd.revents = revents; + QLIST_INSERT_HEAD(ready_list, node, node_ready); +} + #ifdef CONFIG_EPOLL_CREATE1 /* The fd number threshold to switch to epoll */ @@ -XXX,XX +XXX,XX @@ static void aio_epoll_update(AioContext *ctx, AioHandler *node, bool is_new) } } -static int aio_epoll(AioContext *ctx, int64_t timeout) +static int aio_epoll(AioContext *ctx, AioHandlerList *ready_list, + int64_t timeout) { GPollFD pfd = { .fd = ctx->epollfd, @@ -XXX,XX +XXX,XX @@ static int aio_epoll(AioContext *ctx, int64_t timeout) } for (i = 0; i < ret; i++) { int ev = events[i].events; + int revents = (ev & EPOLLIN ? G_IO_IN : 0) | + (ev & EPOLLOUT ? G_IO_OUT : 0) | + (ev & EPOLLHUP ? G_IO_HUP : 0) | + (ev & EPOLLERR ? G_IO_ERR : 0); + node = events[i].data.ptr; - node->pfd.revents = (ev & EPOLLIN ? G_IO_IN : 0) | - (ev & EPOLLOUT ? G_IO_OUT : 0) | - (ev & EPOLLHUP ? G_IO_HUP : 0) | - (ev & EPOLLERR ? G_IO_ERR : 0); + add_ready_handler(ready_list, node, revents); } } out: @@ -XXX,XX +XXX,XX @@ static void aio_epoll_update(AioContext *ctx, AioHandler *node, bool is_new) { } -static int aio_epoll(AioContext *ctx, GPollFD *pfds, - unsigned npfd, int64_t timeout) +static int aio_epoll(AioContext *ctx, AioHandlerList *ready_list, + int64_t timeout) { assert(false); } @@ -XXX,XX +XXX,XX @@ static void aio_free_deleted_handlers(AioContext *ctx) qemu_lockcnt_inc_and_unlock(&ctx->list_lock); } -static bool aio_dispatch_handlers(AioContext *ctx) +static bool aio_dispatch_handler(AioContext *ctx, AioHandler *node) { - AioHandler *node, *tmp; bool progress = false; + int revents; - QLIST_FOREACH_SAFE_RCU(node, &ctx->aio_handlers, node, tmp) { - int revents; + revents = node->pfd.revents & node->pfd.events; + node->pfd.revents = 0; - revents = node->pfd.revents & node->pfd.events; - node->pfd.revents = 0; + if (!QLIST_IS_INSERTED(node, node_deleted) && + (revents & (G_IO_IN | G_IO_HUP | G_IO_ERR)) && + aio_node_check(ctx, node->is_external) && + node->io_read) { + node->io_read(node->opaque); - if (!QLIST_IS_INSERTED(node, node_deleted) && - (revents & (G_IO_IN | G_IO_HUP | G_IO_ERR)) && - aio_node_check(ctx, node->is_external) && - node->io_read) { - node->io_read(node->opaque); - - /* aio_notify() does not count as progress */ - if (node->opaque != &ctx->notifier) { - progress = true; - } - } - if (!QLIST_IS_INSERTED(node, node_deleted) && - (revents & (G_IO_OUT | G_IO_ERR)) && - aio_node_check(ctx, node->is_external) && - node->io_write) { - node->io_write(node->opaque); + /* aio_notify() does not count as progress */ + if (node->opaque != &ctx->notifier) { progress = true; } } + if (!QLIST_IS_INSERTED(node, node_deleted) && + (revents & (G_IO_OUT | G_IO_ERR)) && + aio_node_check(ctx, node->is_external) && + node->io_write) { + node->io_write(node->opaque); + progress = true; + } + + return progress; +} + +/* + * If we have a list of ready handlers then this is more efficient than + * scanning all handlers with aio_dispatch_handlers(). + */ +static bool aio_dispatch_ready_handlers(AioContext *ctx, + AioHandlerList *ready_list) +{ + bool progress = false; + AioHandler *node; + + while ((node = QLIST_FIRST(ready_list))) { + QLIST_SAFE_REMOVE(node, node_ready); + progress = aio_dispatch_handler(ctx, node) || progress; + } + + return progress; +} + +/* Slower than aio_dispatch_ready_handlers() but only used via glib */ +static bool aio_dispatch_handlers(AioContext *ctx) +{ + AioHandler *node, *tmp; + bool progress = false; + + QLIST_FOREACH_SAFE_RCU(node, &ctx->aio_handlers, node, tmp) { + progress = aio_dispatch_handler(ctx, node) || progress; + } return progress; } @@ -XXX,XX +XXX,XX @@ static bool try_poll_mode(AioContext *ctx, int64_t *timeout) bool aio_poll(AioContext *ctx, bool blocking) { + AioHandlerList ready_list = QLIST_HEAD_INITIALIZER(ready_list); AioHandler *node; int i; int ret = 0; @@ -XXX,XX +XXX,XX @@ bool aio_poll(AioContext *ctx, bool blocking) /* wait until next event */ if (aio_epoll_check_poll(ctx, pollfds, npfd, timeout)) { npfd = 0; /* pollfds[] is not being used */ - ret = aio_epoll(ctx, timeout); + ret = aio_epoll(ctx, &ready_list, timeout); } else { ret = qemu_poll_ns(pollfds, npfd, timeout); } @@ -XXX,XX +XXX,XX @@ bool aio_poll(AioContext *ctx, bool blocking) /* if we have any readable fds, dispatch event */ if (ret > 0) { for (i = 0; i < npfd; i++) { - nodes[i]->pfd.revents = pollfds[i].revents; + int revents = pollfds[i].revents; + + if (revents) { + add_ready_handler(&ready_list, nodes[i], revents); + } } } @@ -XXX,XX +XXX,XX @@ bool aio_poll(AioContext *ctx, bool blocking) progress |= aio_bh_poll(ctx); if (ret > 0) { - progress |= aio_dispatch_handlers(ctx); + progress |= aio_dispatch_ready_handlers(ctx, &ready_list); } aio_free_deleted_handlers(ctx); -- 2.24.1 From: Alexander Bulekov <alxndr@bu.edu> Move vl.c to a separate directory, similar to linux-user/ Update the chechpatch and get_maintainer scripts, since they relied on /vl.c for top_of_tree checks. Signed-off-by: Alexander Bulekov <alxndr@bu.edu> Reviewed-by: Darren Kenny <darren.kenny@oracle.com> Message-id: 20200220041118.23264-2-alxndr@bu.edu Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> --- MAINTAINERS | 2 +- Makefile.objs | 2 -- Makefile.target | 1 + scripts/checkpatch.pl | 2 +- scripts/get_maintainer.pl | 3 ++- softmmu/Makefile.objs | 2 ++ vl.c => softmmu/vl.c | 0 7 files changed, 7 insertions(+), 5 deletions(-) create mode 100644 softmmu/Makefile.objs rename vl.c => softmmu/vl.c (100%) diff --git a/MAINTAINERS b/MAINTAINERS index XXXXXXX..XXXXXXX 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -XXX,XX +XXX,XX @@ F: include/qemu/main-loop.h F: include/sysemu/runstate.h F: util/main-loop.c F: util/qemu-timer.c -F: vl.c +F: softmmu/vl.c F: qapi/run-state.json Human Monitor (HMP) diff --git a/Makefile.objs b/Makefile.objs index XXXXXXX..XXXXXXX 100644 --- a/Makefile.objs +++ b/Makefile.objs @@ -XXX,XX +XXX,XX @@ common-obj-y += ui/ common-obj-m += ui/ common-obj-y += dma-helpers.o -common-obj-y += vl.o -vl.o-cflags := $(GPROF_CFLAGS) $(SDL_CFLAGS) common-obj-$(CONFIG_TPM) += tpm.o common-obj-y += backends/ diff --git a/Makefile.target b/Makefile.target index XXXXXXX..XXXXXXX 100644 --- a/Makefile.target +++ b/Makefile.target @@ -XXX,XX +XXX,XX @@ obj-y += qapi/ obj-y += memory.o obj-y += memory_mapping.o obj-y += migration/ram.o +obj-y += softmmu/ LIBS := $(libs_softmmu) $(LIBS) # Hardware support diff --git a/scripts/checkpatch.pl b/scripts/checkpatch.pl index XXXXXXX..XXXXXXX 100755 --- a/scripts/checkpatch.pl +++ b/scripts/checkpatch.pl @@ -XXX,XX +XXX,XX @@ sub top_of_kernel_tree { my @tree_check = ( "COPYING", "MAINTAINERS", "Makefile", "README.rst", "docs", "VERSION", - "vl.c" + "linux-user", "softmmu" ); foreach my $check (@tree_check) { diff --git a/scripts/get_maintainer.pl b/scripts/get_maintainer.pl index XXXXXXX..XXXXXXX 100755 --- a/scripts/get_maintainer.pl +++ b/scripts/get_maintainer.pl @@ -XXX,XX +XXX,XX @@ sub top_of_tree { && (-f "${lk_path}Makefile") && (-d "${lk_path}docs") && (-f "${lk_path}VERSION") - && (-f "${lk_path}vl.c")) { + && (-d "${lk_path}linux-user/") + && (-d "${lk_path}softmmu/")) { return 1; } return 0; diff --git a/softmmu/Makefile.objs b/softmmu/Makefile.objs new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/softmmu/Makefile.objs @@ -XXX,XX +XXX,XX @@ +obj-y += vl.o +vl.o-cflags := $(GPROF_CFLAGS) $(SDL_CFLAGS) diff --git a/vl.c b/softmmu/vl.c similarity index 100% rename from vl.c rename to softmmu/vl.c -- 2.24.1 From: Alexander Bulekov <alxndr@bu.edu> A program might rely on functions implemented in vl.c, but implement its own main(). By placing main into a separate source file, there are no complaints about duplicate main()s when linking against vl.o. For example, the virtual-device fuzzer uses a main() provided by libfuzzer, and needs to perform some initialization before running the softmmu initialization. Now, main simply calls three vl.c functions which handle the guest initialization, main loop and cleanup. Signed-off-by: Alexander Bulekov <alxndr@bu.edu> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Darren Kenny <darren.kenny@oracle.com> Message-id: 20200220041118.23264-3-alxndr@bu.edu Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> --- MAINTAINERS | 1 + Makefile.target | 2 +- include/sysemu/sysemu.h | 4 ++++ softmmu/Makefile.objs | 1 + softmmu/main.c | 53 +++++++++++++++++++++++++++++++++++++++++ softmmu/vl.c | 36 +++++++--------------------- 6 files changed, 69 insertions(+), 28 deletions(-) create mode 100644 softmmu/main.c diff --git a/MAINTAINERS b/MAINTAINERS index XXXXXXX..XXXXXXX 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -XXX,XX +XXX,XX @@ F: include/sysemu/runstate.h F: util/main-loop.c F: util/qemu-timer.c F: softmmu/vl.c +F: softmmu/main.c F: qapi/run-state.json Human Monitor (HMP) diff --git a/Makefile.target b/Makefile.target index XXXXXXX..XXXXXXX 100644 --- a/Makefile.target +++ b/Makefile.target @@ -XXX,XX +XXX,XX @@ endif COMMON_LDADDS = ../libqemuutil.a # build either PROG or PROGW -$(QEMU_PROG_BUILD): $(all-obj-y) $(COMMON_LDADDS) +$(QEMU_PROG_BUILD): $(all-obj-y) $(COMMON_LDADDS) $(softmmu-main-y) $(call LINK, $(filter-out %.mak, $^)) ifdef CONFIG_DARWIN $(call quiet-command,Rez -append $(SRC_PATH)/pc-bios/qemu.rsrc -o $@,"REZ","$(TARGET_DIR)$@") diff --git a/include/sysemu/sysemu.h b/include/sysemu/sysemu.h index XXXXXXX..XXXXXXX 100644 --- a/include/sysemu/sysemu.h +++ b/include/sysemu/sysemu.h @@ -XXX,XX +XXX,XX @@ QemuOpts *qemu_get_machine_opts(void); bool defaults_enabled(void); +void qemu_init(int argc, char **argv, char **envp); +void qemu_main_loop(void); +void qemu_cleanup(void); + extern QemuOptsList qemu_legacy_drive_opts; extern QemuOptsList qemu_common_drive_opts; extern QemuOptsList qemu_drive_opts; diff --git a/softmmu/Makefile.objs b/softmmu/Makefile.objs index XXXXXXX..XXXXXXX 100644 --- a/softmmu/Makefile.objs +++ b/softmmu/Makefile.objs @@ -XXX,XX +XXX,XX @@ +softmmu-main-y = softmmu/main.o obj-y += vl.o vl.o-cflags := $(GPROF_CFLAGS) $(SDL_CFLAGS) diff --git a/softmmu/main.c b/softmmu/main.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/softmmu/main.c @@ -XXX,XX +XXX,XX @@ +/* + * QEMU System Emulator + * + * Copyright (c) 2003-2020 Fabrice Bellard + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to deal + * in the Software without restriction, including without limitation the rights + * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell + * copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, + * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN + * THE SOFTWARE. + */ + +#include "qemu/osdep.h" +#include "qemu-common.h" +#include "sysemu/sysemu.h" + +#ifdef CONFIG_SDL +#if defined(__APPLE__) || defined(main) +#include <SDL.h> +int main(int argc, char **argv) +{ + return qemu_main(argc, argv, NULL); +} +#undef main +#define main qemu_main +#endif +#endif /* CONFIG_SDL */ + +#ifdef CONFIG_COCOA +#undef main +#define main qemu_main +#endif /* CONFIG_COCOA */ + +int main(int argc, char **argv, char **envp) +{ + qemu_init(argc, argv, envp); + qemu_main_loop(); + qemu_cleanup(); + + return 0; +} diff --git a/softmmu/vl.c b/softmmu/vl.c index XXXXXXX..XXXXXXX 100644 --- a/softmmu/vl.c +++ b/softmmu/vl.c @@ -XXX,XX +XXX,XX @@ #include "sysemu/seccomp.h" #include "sysemu/tcg.h" -#ifdef CONFIG_SDL -#if defined(__APPLE__) || defined(main) -#include <SDL.h> -int qemu_main(int argc, char **argv, char **envp); -int main(int argc, char **argv) -{ - return qemu_main(argc, argv, NULL); -} -#undef main -#define main qemu_main -#endif -#endif /* CONFIG_SDL */ - -#ifdef CONFIG_COCOA -#undef main -#define main qemu_main -#endif /* CONFIG_COCOA */ - - #include "qemu/error-report.h" #include "qemu/sockets.h" #include "sysemu/accel.h" @@ -XXX,XX +XXX,XX @@ static bool main_loop_should_exit(void) return false; } -static void main_loop(void) +void qemu_main_loop(void) { #ifdef CONFIG_PROFILER int64_t ti; @@ -XXX,XX +XXX,XX @@ static void configure_accelerators(const char *progname) } } -int main(int argc, char **argv, char **envp) +void qemu_init(int argc, char **argv, char **envp) { int i; int snapshot, linux_boot; @@ -XXX,XX +XXX,XX @@ int main(int argc, char **argv, char **envp) case QEMU_OPTION_watchdog: if (watchdog) { error_report("only one watchdog option may be given"); - return 1; + exit(1); } watchdog = optarg; break; @@ -XXX,XX +XXX,XX @@ int main(int argc, char **argv, char **envp) parse_numa_opts(current_machine); /* do monitor/qmp handling at preconfig state if requested */ - main_loop(); + qemu_main_loop(); audio_init_audiodevs(); @@ -XXX,XX +XXX,XX @@ int main(int argc, char **argv, char **envp) if (vmstate_dump_file) { /* dump and exit */ dump_vmstate_json_to_file(vmstate_dump_file); - return 0; + exit(0); } if (incoming) { @@ -XXX,XX +XXX,XX @@ int main(int argc, char **argv, char **envp) accel_setup_post(current_machine); os_setup_post(); - main_loop(); + return; +} +void qemu_cleanup(void) +{ gdbserver_cleanup(); /* @@ -XXX,XX +XXX,XX @@ int main(int argc, char **argv, char **envp) qemu_chr_cleanup(); user_creatable_cleanup(); /* TODO: unref root container, check all devices are ok */ - - return 0; } -- 2.24.1 From: Alexander Bulekov <alxndr@bu.edu> The virtual-device fuzzer must initialize QOM, prior to running vl:qemu_init, so that it can use the qos_graph to identify the arguments required to initialize a guest for libqos-assisted fuzzing. This change prevents errors when vl:qemu_init tries to (re)initialize the previously initialized QOM module. Signed-off-by: Alexander Bulekov <alxndr@bu.edu> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Darren Kenny <darren.kenny@oracle.com> Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com> Message-id: 20200220041118.23264-4-alxndr@bu.edu Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> --- util/module.c | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/util/module.c b/util/module.c index XXXXXXX..XXXXXXX 100644 --- a/util/module.c +++ b/util/module.c @@ -XXX,XX +XXX,XX @@ typedef struct ModuleEntry typedef QTAILQ_HEAD(, ModuleEntry) ModuleTypeList; static ModuleTypeList init_type_list[MODULE_INIT_MAX]; +static bool modules_init_done[MODULE_INIT_MAX]; static ModuleTypeList dso_init_list; @@ -XXX,XX +XXX,XX @@ void module_call_init(module_init_type type) ModuleTypeList *l; ModuleEntry *e; + if (modules_init_done[type]) { + return; + } + l = find_type(type); QTAILQ_FOREACH(e, l, node) { e->init(); } + + modules_init_done[type] = true; } #ifdef CONFIG_MODULES -- 2.24.1 From: Alexander Bulekov <alxndr@bu.edu> Signed-off-by: Alexander Bulekov <alxndr@bu.edu> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Darren Kenny <darren.kenny@oracle.com> Message-id: 20200220041118.23264-5-alxndr@bu.edu Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> --- include/qemu/module.h | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/include/qemu/module.h b/include/qemu/module.h index XXXXXXX..XXXXXXX 100644 --- a/include/qemu/module.h +++ b/include/qemu/module.h @@ -XXX,XX +XXX,XX @@ typedef enum { MODULE_INIT_TRACE, MODULE_INIT_XEN_BACKEND, MODULE_INIT_LIBQOS, + MODULE_INIT_FUZZ_TARGET, MODULE_INIT_MAX } module_init_type; @@ -XXX,XX +XXX,XX @@ typedef enum { #define xen_backend_init(function) module_init(function, \ MODULE_INIT_XEN_BACKEND) #define libqos_init(function) module_init(function, MODULE_INIT_LIBQOS) - +#define fuzz_target_init(function) module_init(function, \ + MODULE_INIT_FUZZ_TARGET) #define block_module_load_one(lib) module_load_one("block-", lib) #define ui_module_load_one(lib) module_load_one("ui-", lib) #define audio_module_load_one(lib) module_load_one("audio-", lib) -- 2.24.1 From: Alexander Bulekov <alxndr@bu.edu> qtest_server_send is a function pointer specifying the handler used to transmit data to the qtest client. In the standard configuration, this calls the CharBackend handler, but now it is possible for other types of handlers, e.g direct-function calls if the qtest client and server exist within the same process (inproc) Signed-off-by: Alexander Bulekov <alxndr@bu.edu> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Darren Kenny <darren.kenny@oracle.com> Acked-by: Thomas Huth <thuth@redhat.com> Message-id: 20200220041118.23264-6-alxndr@bu.edu Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> --- include/sysemu/qtest.h | 3 +++ qtest.c | 18 ++++++++++++++++-- 2 files changed, 19 insertions(+), 2 deletions(-) diff --git a/include/sysemu/qtest.h b/include/sysemu/qtest.h index XXXXXXX..XXXXXXX 100644 --- a/include/sysemu/qtest.h +++ b/include/sysemu/qtest.h @@ -XXX,XX +XXX,XX @@ bool qtest_driver(void); void qtest_server_init(const char *qtest_chrdev, const char *qtest_log, Error **errp); +void qtest_server_set_send_handler(void (*send)(void *, const char *), + void *opaque); + #endif diff --git a/qtest.c b/qtest.c index XXXXXXX..XXXXXXX 100644 --- a/qtest.c +++ b/qtest.c @@ -XXX,XX +XXX,XX @@ static GString *inbuf; static int irq_levels[MAX_IRQ]; static qemu_timeval start_time; static bool qtest_opened; +static void (*qtest_server_send)(void*, const char*); +static void *qtest_server_send_opaque; #define FMT_timeval "%ld.%06ld" @@ -XXX,XX +XXX,XX @@ static void GCC_FMT_ATTR(1, 2) qtest_log_send(const char *fmt, ...) va_end(ap); } -static void do_qtest_send(CharBackend *chr, const char *str, size_t len) +static void qtest_server_char_be_send(void *opaque, const char *str) { + size_t len = strlen(str); + CharBackend* chr = (CharBackend *)opaque; qemu_chr_fe_write_all(chr, (uint8_t *)str, len); if (qtest_log_fp && qtest_opened) { fprintf(qtest_log_fp, "%s", str); @@ -XXX,XX +XXX,XX @@ static void do_qtest_send(CharBackend *chr, const char *str, size_t len) static void qtest_send(CharBackend *chr, const char *str) { - do_qtest_send(chr, str, strlen(str)); + qtest_server_send(qtest_server_send_opaque, str); } static void GCC_FMT_ATTR(2, 3) qtest_sendf(CharBackend *chr, @@ -XXX,XX +XXX,XX @@ void qtest_server_init(const char *qtest_chrdev, const char *qtest_log, Error ** qemu_chr_fe_set_echo(&qtest_chr, true); inbuf = g_string_new(""); + + if (!qtest_server_send) { + qtest_server_set_send_handler(qtest_server_char_be_send, &qtest_chr); + } +} + +void qtest_server_set_send_handler(void (*send)(void*, const char*), void *opaque) +{ + qtest_server_send = send; + qtest_server_send_opaque = opaque; } bool qtest_driver(void) -- 2.24.1 From: Alexander Bulekov <alxndr@bu.edu> This makes it simple to swap the transport functions for qtest commands to and from the qtest client. For example, now it is possible to directly pass qtest commands to a server handler that exists within the same process, without the standard way of writing to a file descriptor. Signed-off-by: Alexander Bulekov <alxndr@bu.edu> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Darren Kenny <darren.kenny@oracle.com> Message-id: 20200220041118.23264-7-alxndr@bu.edu Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> --- tests/qtest/libqtest.c | 48 ++++++++++++++++++++++++++++++++++-------- 1 file changed, 39 insertions(+), 9 deletions(-) diff --git a/tests/qtest/libqtest.c b/tests/qtest/libqtest.c index XXXXXXX..XXXXXXX 100644 --- a/tests/qtest/libqtest.c +++ b/tests/qtest/libqtest.c @@ -XXX,XX +XXX,XX @@ #define SOCKET_TIMEOUT 50 #define SOCKET_MAX_FDS 16 + +typedef void (*QTestSendFn)(QTestState *s, const char *buf); +typedef GString* (*QTestRecvFn)(QTestState *); + +typedef struct QTestClientTransportOps { + QTestSendFn send; /* for sending qtest commands */ + QTestRecvFn recv_line; /* for receiving qtest command responses */ +} QTestTransportOps; + struct QTestState { int fd; @@ -XXX,XX +XXX,XX @@ struct QTestState bool big_endian; bool irq_level[MAX_IRQ]; GString *rx; + QTestTransportOps ops; }; static GHookList abrt_hooks; @@ -XXX,XX +XXX,XX @@ static struct sigaction sigact_old; static int qtest_query_target_endianness(QTestState *s); +static void qtest_client_socket_send(QTestState*, const char *buf); +static void socket_send(int fd, const char *buf, size_t size); + +static GString *qtest_client_socket_recv_line(QTestState *); + +static void qtest_client_set_tx_handler(QTestState *s, QTestSendFn send); +static void qtest_client_set_rx_handler(QTestState *s, QTestRecvFn recv); + static int init_socket(const char *socket_path) { struct sockaddr_un addr; @@ -XXX,XX +XXX,XX @@ QTestState *qtest_init_without_qmp_handshake(const char *extra_args) sock = init_socket(socket_path); qmpsock = init_socket(qmp_socket_path); + qtest_client_set_rx_handler(s, qtest_client_socket_recv_line); + qtest_client_set_tx_handler(s, qtest_client_socket_send); + qtest_add_abrt_handler(kill_qemu_hook_func, s); command = g_strdup_printf("exec %s " @@ -XXX,XX +XXX,XX @@ static void socket_send(int fd, const char *buf, size_t size) } } -static void socket_sendf(int fd, const char *fmt, va_list ap) +static void qtest_client_socket_send(QTestState *s, const char *buf) { - gchar *str = g_strdup_vprintf(fmt, ap); - size_t size = strlen(str); - - socket_send(fd, str, size); - g_free(str); + socket_send(s->fd, buf, strlen(buf)); } static void GCC_FMT_ATTR(2, 3) qtest_sendf(QTestState *s, const char *fmt, ...) @@ -XXX,XX +XXX,XX @@ static void GCC_FMT_ATTR(2, 3) qtest_sendf(QTestState *s, const char *fmt, ...) va_list ap; va_start(ap, fmt); - socket_sendf(s->fd, fmt, ap); + gchar *str = g_strdup_vprintf(fmt, ap); va_end(ap); + + s->ops.send(s, str); + g_free(str); } /* Sends a message and file descriptors to the socket. @@ -XXX,XX +XXX,XX @@ static void socket_send_fds(int socket_fd, int *fds, size_t fds_num, g_assert_cmpint(ret, >, 0); } -static GString *qtest_recv_line(QTestState *s) +static GString *qtest_client_socket_recv_line(QTestState *s) { GString *line; size_t offset; @@ -XXX,XX +XXX,XX @@ static gchar **qtest_rsp(QTestState *s, int expected_args) int i; redo: - line = qtest_recv_line(s); + line = s->ops.recv_line(s); words = g_strsplit(line->str, " ", 0); g_string_free(line, TRUE); @@ -XXX,XX +XXX,XX @@ void qmp_assert_error_class(QDict *rsp, const char *class) qobject_unref(rsp); } + +static void qtest_client_set_tx_handler(QTestState *s, + QTestSendFn send) +{ + s->ops.send = send; +} +static void qtest_client_set_rx_handler(QTestState *s, QTestRecvFn recv) +{ + s->ops.recv_line = recv; +} -- 2.24.1 From: Alexander Bulekov <alxndr@bu.edu> When using qtest "in-process" communication, qtest_sendf directly calls a function in the server (qtest.c). Previously, bufwrite used socket_send, which bypasses the TransportOps enabling the call into qtest.c. This change replaces the socket_send calls with ops->send, maintaining the benefits of the direct socket_send call, while adding support for in-process qtest calls. Signed-off-by: Alexander Bulekov <alxndr@bu.edu> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Darren Kenny <darren.kenny@oracle.com> Message-id: 20200220041118.23264-8-alxndr@bu.edu Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> --- tests/qtest/libqtest.c | 71 ++++++++++++++++++++++++++++++++++++++++-- tests/qtest/libqtest.h | 4 +++ 2 files changed, 73 insertions(+), 2 deletions(-) diff --git a/tests/qtest/libqtest.c b/tests/qtest/libqtest.c index XXXXXXX..XXXXXXX 100644 --- a/tests/qtest/libqtest.c +++ b/tests/qtest/libqtest.c @@ -XXX,XX +XXX,XX @@ typedef void (*QTestSendFn)(QTestState *s, const char *buf); +typedef void (*ExternalSendFn)(void *s, const char *buf); typedef GString* (*QTestRecvFn)(QTestState *); typedef struct QTestClientTransportOps { QTestSendFn send; /* for sending qtest commands */ + + /* + * use external_send to send qtest command strings through functions which + * do not accept a QTestState as the first parameter. + */ + ExternalSendFn external_send; + QTestRecvFn recv_line; /* for receiving qtest command responses */ } QTestTransportOps; @@ -XXX,XX +XXX,XX @@ void qtest_bufwrite(QTestState *s, uint64_t addr, const void *data, size_t size) bdata = g_base64_encode(data, size); qtest_sendf(s, "b64write 0x%" PRIx64 " 0x%zx ", addr, size); - socket_send(s->fd, bdata, strlen(bdata)); - socket_send(s->fd, "\n", 1); + s->ops.send(s, bdata); + s->ops.send(s, "\n"); qtest_rsp(s, 0); g_free(bdata); } @@ -XXX,XX +XXX,XX @@ static void qtest_client_set_rx_handler(QTestState *s, QTestRecvFn recv) { s->ops.recv_line = recv; } +/* A type-safe wrapper for s->send() */ +static void send_wrapper(QTestState *s, const char *buf) +{ + s->ops.external_send(s, buf); +} + +static GString *qtest_client_inproc_recv_line(QTestState *s) +{ + GString *line; + size_t offset; + char *eol; + + eol = strchr(s->rx->str, '\n'); + offset = eol - s->rx->str; + line = g_string_new_len(s->rx->str, offset); + g_string_erase(s->rx, 0, offset + 1); + return line; +} + +QTestState *qtest_inproc_init(QTestState **s, bool log, const char* arch, + void (*send)(void*, const char*)) +{ + QTestState *qts; + qts = g_new0(QTestState, 1); + *s = qts; /* Expose qts early on, since the query endianness relies on it */ + qts->wstatus = 0; + for (int i = 0; i < MAX_IRQ; i++) { + qts->irq_level[i] = false; + } + + qtest_client_set_rx_handler(qts, qtest_client_inproc_recv_line); + + /* send() may not have a matching protoype, so use a type-safe wrapper */ + qts->ops.external_send = send; + qtest_client_set_tx_handler(qts, send_wrapper); + + qts->big_endian = qtest_query_target_endianness(qts); + + /* + * Set a dummy path for QTEST_QEMU_BINARY. Doesn't need to exist, but this + * way, qtest_get_arch works for inproc qtest. + */ + gchar *bin_path = g_strconcat("/qemu-system-", arch, NULL); + setenv("QTEST_QEMU_BINARY", bin_path, 0); + g_free(bin_path); + + return qts; +} + +void qtest_client_inproc_recv(void *opaque, const char *str) +{ + QTestState *qts = *(QTestState **)opaque; + + if (!qts->rx) { + qts->rx = g_string_new(NULL); + } + g_string_append(qts->rx, str); + return; +} diff --git a/tests/qtest/libqtest.h b/tests/qtest/libqtest.h index XXXXXXX..XXXXXXX 100644 --- a/tests/qtest/libqtest.h +++ b/tests/qtest/libqtest.h @@ -XXX,XX +XXX,XX @@ bool qtest_probe_child(QTestState *s); */ void qtest_set_expected_status(QTestState *s, int status); +QTestState *qtest_inproc_init(QTestState **s, bool log, const char* arch, + void (*send)(void*, const char*)); + +void qtest_client_inproc_recv(void *opaque, const char *str); #endif -- 2.24.1 From: Alexander Bulekov <alxndr@bu.edu> The handler allows a qtest client to send commands to the server by directly calling a function, rather than using a file/CharBackend Signed-off-by: Alexander Bulekov <alxndr@bu.edu> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Darren Kenny <darren.kenny@oracle.com> Message-id: 20200220041118.23264-9-alxndr@bu.edu Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> --- include/sysemu/qtest.h | 1 + qtest.c | 13 +++++++++++++ 2 files changed, 14 insertions(+) diff --git a/include/sysemu/qtest.h b/include/sysemu/qtest.h index XXXXXXX..XXXXXXX 100644 --- a/include/sysemu/qtest.h +++ b/include/sysemu/qtest.h @@ -XXX,XX +XXX,XX @@ void qtest_server_init(const char *qtest_chrdev, const char *qtest_log, Error ** void qtest_server_set_send_handler(void (*send)(void *, const char *), void *opaque); +void qtest_server_inproc_recv(void *opaque, const char *buf); #endif diff --git a/qtest.c b/qtest.c index XXXXXXX..XXXXXXX 100644 --- a/qtest.c +++ b/qtest.c @@ -XXX,XX +XXX,XX @@ bool qtest_driver(void) { return qtest_chr.chr != NULL; } + +void qtest_server_inproc_recv(void *dummy, const char *buf) +{ + static GString *gstr; + if (!gstr) { + gstr = g_string_new(NULL); + } + g_string_append(gstr, buf); + if (gstr->str[gstr->len - 1] == '\n') { + qtest_process_inbuf(NULL, gstr); + g_string_truncate(gstr, 0); + } +} -- 2.24.1 From: Alexander Bulekov <alxndr@bu.edu> The names i2c_send and i2c_recv collide with functions defined in hw/i2c/core.c. This causes an error when linking against libqos and softmmu simultaneously (for example when using qtest inproc). Rename the libqos functions to avoid this. Signed-off-by: Alexander Bulekov <alxndr@bu.edu> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Darren Kenny <darren.kenny@oracle.com> Acked-by: Thomas Huth <thuth@redhat.com> Message-id: 20200220041118.23264-10-alxndr@bu.edu Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> --- tests/qtest/libqos/i2c.c | 10 +++++----- tests/qtest/libqos/i2c.h | 4 ++-- tests/qtest/pca9552-test.c | 10 +++++----- 3 files changed, 12 insertions(+), 12 deletions(-) diff --git a/tests/qtest/libqos/i2c.c b/tests/qtest/libqos/i2c.c index XXXXXXX..XXXXXXX 100644 --- a/tests/qtest/libqos/i2c.c +++ b/tests/qtest/libqos/i2c.c @@ -XXX,XX +XXX,XX @@ #include "libqos/i2c.h" #include "libqtest.h" -void i2c_send(QI2CDevice *i2cdev, const uint8_t *buf, uint16_t len) +void qi2c_send(QI2CDevice *i2cdev, const uint8_t *buf, uint16_t len) { i2cdev->bus->send(i2cdev->bus, i2cdev->addr, buf, len); } -void i2c_recv(QI2CDevice *i2cdev, uint8_t *buf, uint16_t len) +void qi2c_recv(QI2CDevice *i2cdev, uint8_t *buf, uint16_t len) { i2cdev->bus->recv(i2cdev->bus, i2cdev->addr, buf, len); } @@ -XXX,XX +XXX,XX @@ void i2c_recv(QI2CDevice *i2cdev, uint8_t *buf, uint16_t len) void i2c_read_block(QI2CDevice *i2cdev, uint8_t reg, uint8_t *buf, uint16_t len) { - i2c_send(i2cdev, &reg, 1); - i2c_recv(i2cdev, buf, len); + qi2c_send(i2cdev, &reg, 1); + qi2c_recv(i2cdev, buf, len); } void i2c_write_block(QI2CDevice *i2cdev, uint8_t reg, @@ -XXX,XX +XXX,XX @@ void i2c_write_block(QI2CDevice *i2cdev, uint8_t reg, uint8_t *cmd = g_malloc(len + 1); cmd[0] = reg; memcpy(&cmd[1], buf, len); - i2c_send(i2cdev, cmd, len + 1); + qi2c_send(i2cdev, cmd, len + 1); g_free(cmd); } diff --git a/tests/qtest/libqos/i2c.h b/tests/qtest/libqos/i2c.h index XXXXXXX..XXXXXXX 100644 --- a/tests/qtest/libqos/i2c.h +++ b/tests/qtest/libqos/i2c.h @@ -XXX,XX +XXX,XX @@ struct QI2CDevice { void *i2c_device_create(void *i2c_bus, QGuestAllocator *alloc, void *addr); void add_qi2c_address(QOSGraphEdgeOptions *opts, QI2CAddress *addr); -void i2c_send(QI2CDevice *dev, const uint8_t *buf, uint16_t len); -void i2c_recv(QI2CDevice *dev, uint8_t *buf, uint16_t len); +void qi2c_send(QI2CDevice *dev, const uint8_t *buf, uint16_t len); +void qi2c_recv(QI2CDevice *dev, uint8_t *buf, uint16_t len); void i2c_read_block(QI2CDevice *dev, uint8_t reg, uint8_t *buf, uint16_t len); diff --git a/tests/qtest/pca9552-test.c b/tests/qtest/pca9552-test.c index XXXXXXX..XXXXXXX 100644 --- a/tests/qtest/pca9552-test.c +++ b/tests/qtest/pca9552-test.c @@ -XXX,XX +XXX,XX @@ static void receive_autoinc(void *obj, void *data, QGuestAllocator *alloc) pca9552_init(i2cdev); - i2c_send(i2cdev, &reg, 1); + qi2c_send(i2cdev, &reg, 1); /* PCA9552_LS0 */ - i2c_recv(i2cdev, &resp, 1); + qi2c_recv(i2cdev, &resp, 1); g_assert_cmphex(resp, ==, 0x54); /* PCA9552_LS1 */ - i2c_recv(i2cdev, &resp, 1); + qi2c_recv(i2cdev, &resp, 1); g_assert_cmphex(resp, ==, 0x55); /* PCA9552_LS2 */ - i2c_recv(i2cdev, &resp, 1); + qi2c_recv(i2cdev, &resp, 1); g_assert_cmphex(resp, ==, 0x55); /* PCA9552_LS3 */ - i2c_recv(i2cdev, &resp, 1); + qi2c_recv(i2cdev, &resp, 1); g_assert_cmphex(resp, ==, 0x54); } -- 2.24.1 From: Alexander Bulekov <alxndr@bu.edu> Most qos-related objects were specified in the qos-test-obj-y variable. qos-test-obj-y also included qos-test.o which defines a main(). This made it difficult to repurpose qos-test-obj-y to link anything beside tests/qos-test against libqos. This change separates objects that are libqos-specific and ones that are qos-test specific into different variables. Signed-off-by: Alexander Bulekov <alxndr@bu.edu> Reviewed-by: Darren Kenny <darren.kenny@oracle.com> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com> Message-id: 20200220041118.23264-11-alxndr@bu.edu Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> --- tests/qtest/Makefile.include | 71 ++++++++++++++++++------------------ 1 file changed, 36 insertions(+), 35 deletions(-) diff --git a/tests/qtest/Makefile.include b/tests/qtest/Makefile.include index XXXXXXX..XXXXXXX 100644 --- a/tests/qtest/Makefile.include +++ b/tests/qtest/Makefile.include @@ -XXX,XX +XXX,XX @@ check-qtest-s390x-y += migration-test # libqos / qgraph : libqgraph-obj-y = tests/qtest/libqos/qgraph.o -libqos-obj-y = $(libqgraph-obj-y) tests/qtest/libqos/pci.o tests/qtest/libqos/fw_cfg.o -libqos-obj-y += tests/qtest/libqos/malloc.o -libqos-obj-y += tests/qtest/libqos/libqos.o -libqos-spapr-obj-y = $(libqos-obj-y) tests/qtest/libqos/malloc-spapr.o +libqos-core-obj-y = $(libqgraph-obj-y) tests/qtest/libqos/pci.o tests/qtest/libqos/fw_cfg.o +libqos-core-obj-y += tests/qtest/libqos/malloc.o +libqos-core-obj-y += tests/qtest/libqos/libqos.o +libqos-spapr-obj-y = $(libqos-core-obj-y) tests/qtest/libqos/malloc-spapr.o libqos-spapr-obj-y += tests/qtest/libqos/libqos-spapr.o libqos-spapr-obj-y += tests/qtest/libqos/rtas.o libqos-spapr-obj-y += tests/qtest/libqos/pci-spapr.o -libqos-pc-obj-y = $(libqos-obj-y) tests/qtest/libqos/pci-pc.o +libqos-pc-obj-y = $(libqos-core-obj-y) tests/qtest/libqos/pci-pc.o libqos-pc-obj-y += tests/qtest/libqos/malloc-pc.o tests/qtest/libqos/libqos-pc.o libqos-pc-obj-y += tests/qtest/libqos/ahci.o libqos-usb-obj-y = $(libqos-spapr-obj-y) $(libqos-pc-obj-y) tests/qtest/libqos/usb.o # qos devices: -qos-test-obj-y = tests/qtest/qos-test.o $(libqgraph-obj-y) -qos-test-obj-y += $(libqos-pc-obj-y) $(libqos-spapr-obj-y) -qos-test-obj-y += tests/qtest/libqos/e1000e.o -qos-test-obj-y += tests/qtest/libqos/i2c.o -qos-test-obj-y += tests/qtest/libqos/i2c-imx.o -qos-test-obj-y += tests/qtest/libqos/i2c-omap.o -qos-test-obj-y += tests/qtest/libqos/sdhci.o -qos-test-obj-y += tests/qtest/libqos/tpci200.o -qos-test-obj-y += tests/qtest/libqos/virtio.o -qos-test-obj-$(CONFIG_VIRTFS) += tests/qtest/libqos/virtio-9p.o -qos-test-obj-y += tests/qtest/libqos/virtio-balloon.o -qos-test-obj-y += tests/qtest/libqos/virtio-blk.o -qos-test-obj-y += tests/qtest/libqos/virtio-mmio.o -qos-test-obj-y += tests/qtest/libqos/virtio-net.o -qos-test-obj-y += tests/qtest/libqos/virtio-pci.o -qos-test-obj-y += tests/qtest/libqos/virtio-pci-modern.o -qos-test-obj-y += tests/qtest/libqos/virtio-rng.o -qos-test-obj-y += tests/qtest/libqos/virtio-scsi.o -qos-test-obj-y += tests/qtest/libqos/virtio-serial.o +libqos-obj-y = $(libqgraph-obj-y) +libqos-obj-y += $(libqos-pc-obj-y) $(libqos-spapr-obj-y) +libqos-obj-y += tests/qtest/libqos/e1000e.o +libqos-obj-y += tests/qtest/libqos/i2c.o +libqos-obj-y += tests/qtest/libqos/i2c-imx.o +libqos-obj-y += tests/qtest/libqos/i2c-omap.o +libqos-obj-y += tests/qtest/libqos/sdhci.o +libqos-obj-y += tests/qtest/libqos/tpci200.o +libqos-obj-y += tests/qtest/libqos/virtio.o +libqos-obj-$(CONFIG_VIRTFS) += tests/qtest/libqos/virtio-9p.o +libqos-obj-y += tests/qtest/libqos/virtio-balloon.o +libqos-obj-y += tests/qtest/libqos/virtio-blk.o +libqos-obj-y += tests/qtest/libqos/virtio-mmio.o +libqos-obj-y += tests/qtest/libqos/virtio-net.o +libqos-obj-y += tests/qtest/libqos/virtio-pci.o +libqos-obj-y += tests/qtest/libqos/virtio-pci-modern.o +libqos-obj-y += tests/qtest/libqos/virtio-rng.o +libqos-obj-y += tests/qtest/libqos/virtio-scsi.o +libqos-obj-y += tests/qtest/libqos/virtio-serial.o # qos machines: -qos-test-obj-y += tests/qtest/libqos/aarch64-xlnx-zcu102-machine.o -qos-test-obj-y += tests/qtest/libqos/arm-imx25-pdk-machine.o -qos-test-obj-y += tests/qtest/libqos/arm-n800-machine.o -qos-test-obj-y += tests/qtest/libqos/arm-raspi2-machine.o -qos-test-obj-y += tests/qtest/libqos/arm-sabrelite-machine.o -qos-test-obj-y += tests/qtest/libqos/arm-smdkc210-machine.o -qos-test-obj-y += tests/qtest/libqos/arm-virt-machine.o -qos-test-obj-y += tests/qtest/libqos/arm-xilinx-zynq-a9-machine.o -qos-test-obj-y += tests/qtest/libqos/ppc64_pseries-machine.o -qos-test-obj-y += tests/qtest/libqos/x86_64_pc-machine.o +libqos-obj-y += tests/qtest/libqos/aarch64-xlnx-zcu102-machine.o +libqos-obj-y += tests/qtest/libqos/arm-imx25-pdk-machine.o +libqos-obj-y += tests/qtest/libqos/arm-n800-machine.o +libqos-obj-y += tests/qtest/libqos/arm-raspi2-machine.o +libqos-obj-y += tests/qtest/libqos/arm-sabrelite-machine.o +libqos-obj-y += tests/qtest/libqos/arm-smdkc210-machine.o +libqos-obj-y += tests/qtest/libqos/arm-virt-machine.o +libqos-obj-y += tests/qtest/libqos/arm-xilinx-zynq-a9-machine.o +libqos-obj-y += tests/qtest/libqos/ppc64_pseries-machine.o +libqos-obj-y += tests/qtest/libqos/x86_64_pc-machine.o # qos tests: +qos-test-obj-y += tests/qtest/qos-test.o qos-test-obj-y += tests/qtest/ac97-test.o qos-test-obj-y += tests/qtest/ds1338-test.o qos-test-obj-y += tests/qtest/e1000-test.o @@ -XXX,XX +XXX,XX @@ check-unit-y += tests/test-qgraph$(EXESUF) tests/test-qgraph$(EXESUF): tests/test-qgraph.o $(libqgraph-obj-y) check-qtest-generic-y += qos-test -tests/qtest/qos-test$(EXESUF): $(qos-test-obj-y) +tests/qtest/qos-test$(EXESUF): $(qos-test-obj-y) $(libqos-obj-y) # QTest dependencies: tests/qtest/qmp-test$(EXESUF): tests/qtest/qmp-test.o -- 2.24.1 From: Alexander Bulekov <alxndr@bu.edu> The moved functions are not specific to qos-test and might be useful elsewhere. For example the virtual-device fuzzer makes use of them for qos-assisted fuzz-targets. Signed-off-by: Alexander Bulekov <alxndr@bu.edu> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com> Reviewed-by: Darren Kenny <darren.kenny@oracle.com> Message-id: 20200220041118.23264-12-alxndr@bu.edu Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> --- tests/qtest/Makefile.include | 1 + tests/qtest/libqos/qos_external.c | 168 ++++++++++++++++++++++++++++++ tests/qtest/libqos/qos_external.h | 28 +++++ tests/qtest/qos-test.c | 132 +---------------------- 4 files changed, 198 insertions(+), 131 deletions(-) create mode 100644 tests/qtest/libqos/qos_external.c create mode 100644 tests/qtest/libqos/qos_external.h diff --git a/tests/qtest/Makefile.include b/tests/qtest/Makefile.include index XXXXXXX..XXXXXXX 100644 --- a/tests/qtest/Makefile.include +++ b/tests/qtest/Makefile.include @@ -XXX,XX +XXX,XX @@ libqos-usb-obj-y = $(libqos-spapr-obj-y) $(libqos-pc-obj-y) tests/qtest/libqos/u # qos devices: libqos-obj-y = $(libqgraph-obj-y) libqos-obj-y += $(libqos-pc-obj-y) $(libqos-spapr-obj-y) +libqos-obj-y += tests/qtest/libqos/qos_external.o libqos-obj-y += tests/qtest/libqos/e1000e.o libqos-obj-y += tests/qtest/libqos/i2c.o libqos-obj-y += tests/qtest/libqos/i2c-imx.o diff --git a/tests/qtest/libqos/qos_external.c b/tests/qtest/libqos/qos_external.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/tests/qtest/libqos/qos_external.c @@ -XXX,XX +XXX,XX @@ +/* + * libqos driver framework + * + * Copyright (c) 2018 Emanuele Giuseppe Esposito <e.emanuelegiuseppe@gmail.com> + * + * This library is free software; you can redistribute it and/or + * modify it under the terms of the GNU Lesser General Public + * License version 2 as published by the Free Software Foundation. + * + * This library is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public + * License along with this library; if not, see <http://www.gnu.org/licenses/> + */ + +#include "qemu/osdep.h" +#include <getopt.h> +#include "libqtest.h" +#include "qapi/qmp/qdict.h" +#include "qapi/qmp/qbool.h" +#include "qapi/qmp/qstring.h" +#include "qemu/module.h" +#include "qapi/qmp/qlist.h" +#include "libqos/malloc.h" +#include "libqos/qgraph.h" +#include "libqos/qgraph_internal.h" +#include "libqos/qos_external.h" + + + +void apply_to_node(const char *name, bool is_machine, bool is_abstract) +{ + char *machine_name = NULL; + if (is_machine) { + const char *arch = qtest_get_arch(); + machine_name = g_strconcat(arch, "/", name, NULL); + name = machine_name; + } + qos_graph_node_set_availability(name, true); + if (is_abstract) { + qos_delete_cmd_line(name); + } + g_free(machine_name); +} + +/** + * apply_to_qlist(): using QMP queries QEMU for a list of + * machines and devices available, and sets the respective node + * as true. If a node is found, also all its produced and contained + * child are marked available. + * + * See qos_graph_node_set_availability() for more info + */ +void apply_to_qlist(QList *list, bool is_machine) +{ + const QListEntry *p; + const char *name; + bool abstract; + QDict *minfo; + QObject *qobj; + QString *qstr; + QBool *qbool; + + for (p = qlist_first(list); p; p = qlist_next(p)) { + minfo = qobject_to(QDict, qlist_entry_obj(p)); + qobj = qdict_get(minfo, "name"); + qstr = qobject_to(QString, qobj); + name = qstring_get_str(qstr); + + qobj = qdict_get(minfo, "abstract"); + if (qobj) { + qbool = qobject_to(QBool, qobj); + abstract = qbool_get_bool(qbool); + } else { + abstract = false; + } + + apply_to_node(name, is_machine, abstract); + qobj = qdict_get(minfo, "alias"); + if (qobj) { + qstr = qobject_to(QString, qobj); + name = qstring_get_str(qstr); + apply_to_node(name, is_machine, abstract); + } + } +} + +QGuestAllocator *get_machine_allocator(QOSGraphObject *obj) +{ + return obj->get_driver(obj, "memory"); +} + +/** + * allocate_objects(): given an array of nodes @arg, + * walks the path invoking all constructors and + * passing the corresponding parameter in order to + * continue the objects allocation. + * Once the test is reached, return the object it consumes. + * + * Since the machine and QEDGE_CONSUMED_BY nodes allocate + * memory in the constructor, g_test_queue_destroy is used so + * that after execution they can be safely free'd. (The test's + * ->before callback is also welcome to use g_test_queue_destroy). + * + * Note: as specified in walk_path() too, @arg is an array of + * char *, where arg[0] is a pointer to the command line + * string that will be used to properly start QEMU when executing + * the test, and the remaining elements represent the actual objects + * that will be allocated. + */ +void *allocate_objects(QTestState *qts, char **path, QGuestAllocator **p_alloc) +{ + int current = 0; + QGuestAllocator *alloc; + QOSGraphObject *parent = NULL; + QOSGraphEdge *edge; + QOSGraphNode *node; + void *edge_arg; + void *obj; + + node = qos_graph_get_node(path[current]); + g_assert(node->type == QNODE_MACHINE); + + obj = qos_machine_new(node, qts); + qos_object_queue_destroy(obj); + + alloc = get_machine_allocator(obj); + if (p_alloc) { + *p_alloc = alloc; + } + + for (;;) { + if (node->type != QNODE_INTERFACE) { + qos_object_start_hw(obj); + parent = obj; + } + + /* follow edge and get object for next node constructor */ + current++; + edge = qos_graph_get_edge(path[current - 1], path[current]); + node = qos_graph_get_node(path[current]); + + if (node->type == QNODE_TEST) { + g_assert(qos_graph_edge_get_type(edge) == QEDGE_CONSUMED_BY); + return obj; + } + + switch (qos_graph_edge_get_type(edge)) { + case QEDGE_PRODUCES: + obj = parent->get_driver(parent, path[current]); + break; + + case QEDGE_CONSUMED_BY: + edge_arg = qos_graph_edge_get_arg(edge); + obj = qos_driver_new(node, obj, alloc, edge_arg); + qos_object_queue_destroy(obj); + break; + + case QEDGE_CONTAINS: + obj = parent->get_device(parent, path[current]); + break; + } + } +} + diff --git a/tests/qtest/libqos/qos_external.h b/tests/qtest/libqos/qos_external.h new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/tests/qtest/libqos/qos_external.h @@ -XXX,XX +XXX,XX @@ +/* + * libqos driver framework + * + * Copyright (c) 2018 Emanuele Giuseppe Esposito <e.emanuelegiuseppe@gmail.com> + * + * This library is free software; you can redistribute it and/or + * modify it under the terms of the GNU Lesser General Public + * License version 2 as published by the Free Software Foundation. + * + * This library is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public + * License along with this library; if not, see <http://www.gnu.org/licenses/> + */ + +#ifndef QOS_EXTERNAL_H +#define QOS_EXTERNAL_H +#include "libqos/qgraph.h" + +void apply_to_node(const char *name, bool is_machine, bool is_abstract); +void apply_to_qlist(QList *list, bool is_machine); +QGuestAllocator *get_machine_allocator(QOSGraphObject *obj); +void *allocate_objects(QTestState *qts, char **path, QGuestAllocator **p_alloc); + +#endif diff --git a/tests/qtest/qos-test.c b/tests/qtest/qos-test.c index XXXXXXX..XXXXXXX 100644 --- a/tests/qtest/qos-test.c +++ b/tests/qtest/qos-test.c @@ -XXX,XX +XXX,XX @@ #include "libqos/malloc.h" #include "libqos/qgraph.h" #include "libqos/qgraph_internal.h" +#include "libqos/qos_external.h" static char *old_path; -static void apply_to_node(const char *name, bool is_machine, bool is_abstract) -{ - char *machine_name = NULL; - if (is_machine) { - const char *arch = qtest_get_arch(); - machine_name = g_strconcat(arch, "/", name, NULL); - name = machine_name; - } - qos_graph_node_set_availability(name, true); - if (is_abstract) { - qos_delete_cmd_line(name); - } - g_free(machine_name); -} -/** - * apply_to_qlist(): using QMP queries QEMU for a list of - * machines and devices available, and sets the respective node - * as true. If a node is found, also all its produced and contained - * child are marked available. - * - * See qos_graph_node_set_availability() for more info - */ -static void apply_to_qlist(QList *list, bool is_machine) -{ - const QListEntry *p; - const char *name; - bool abstract; - QDict *minfo; - QObject *qobj; - QString *qstr; - QBool *qbool; - - for (p = qlist_first(list); p; p = qlist_next(p)) { - minfo = qobject_to(QDict, qlist_entry_obj(p)); - qobj = qdict_get(minfo, "name"); - qstr = qobject_to(QString, qobj); - name = qstring_get_str(qstr); - - qobj = qdict_get(minfo, "abstract"); - if (qobj) { - qbool = qobject_to(QBool, qobj); - abstract = qbool_get_bool(qbool); - } else { - abstract = false; - } - - apply_to_node(name, is_machine, abstract); - qobj = qdict_get(minfo, "alias"); - if (qobj) { - qstr = qobject_to(QString, qobj); - name = qstring_get_str(qstr); - apply_to_node(name, is_machine, abstract); - } - } -} /** * qos_set_machines_devices_available(): sets availability of qgraph @@ -XXX,XX +XXX,XX @@ static void qos_set_machines_devices_available(void) qobject_unref(response); } -static QGuestAllocator *get_machine_allocator(QOSGraphObject *obj) -{ - return obj->get_driver(obj, "memory"); -} static void restart_qemu_or_continue(char *path) { @@ -XXX,XX +XXX,XX @@ void qos_invalidate_command_line(void) old_path = NULL; } -/** - * allocate_objects(): given an array of nodes @arg, - * walks the path invoking all constructors and - * passing the corresponding parameter in order to - * continue the objects allocation. - * Once the test is reached, return the object it consumes. - * - * Since the machine and QEDGE_CONSUMED_BY nodes allocate - * memory in the constructor, g_test_queue_destroy is used so - * that after execution they can be safely free'd. (The test's - * ->before callback is also welcome to use g_test_queue_destroy). - * - * Note: as specified in walk_path() too, @arg is an array of - * char *, where arg[0] is a pointer to the command line - * string that will be used to properly start QEMU when executing - * the test, and the remaining elements represent the actual objects - * that will be allocated. - */ -static void *allocate_objects(QTestState *qts, char **path, QGuestAllocator **p_alloc) -{ - int current = 0; - QGuestAllocator *alloc; - QOSGraphObject *parent = NULL; - QOSGraphEdge *edge; - QOSGraphNode *node; - void *edge_arg; - void *obj; - - node = qos_graph_get_node(path[current]); - g_assert(node->type == QNODE_MACHINE); - - obj = qos_machine_new(node, qts); - qos_object_queue_destroy(obj); - - alloc = get_machine_allocator(obj); - if (p_alloc) { - *p_alloc = alloc; - } - - for (;;) { - if (node->type != QNODE_INTERFACE) { - qos_object_start_hw(obj); - parent = obj; - } - - /* follow edge and get object for next node constructor */ - current++; - edge = qos_graph_get_edge(path[current - 1], path[current]); - node = qos_graph_get_node(path[current]); - - if (node->type == QNODE_TEST) { - g_assert(qos_graph_edge_get_type(edge) == QEDGE_CONSUMED_BY); - return obj; - } - - switch (qos_graph_edge_get_type(edge)) { - case QEDGE_PRODUCES: - obj = parent->get_driver(parent, path[current]); - break; - - case QEDGE_CONSUMED_BY: - edge_arg = qos_graph_edge_get_arg(edge); - obj = qos_driver_new(node, obj, alloc, edge_arg); - qos_object_queue_destroy(obj); - break; - - case QEDGE_CONTAINS: - obj = parent->get_device(parent, path[current]); - break; - } - } -} /* The argument to run_one_test, which is the test function that is registered * with GTest, is a vector of strings. The first item is the initial command -- 2.24.1 From: Alexander Bulekov <alxndr@bu.edu> tests/fuzz/fuzz.c serves as the entry point for the virtual-device fuzzer. Namely, libfuzzer invokes the LLVMFuzzerInitialize and LLVMFuzzerTestOneInput functions, both of which are defined in this file. This change adds a "FuzzTarget" struct, along with the fuzz_add_target function, which should be used to define new fuzz targets. Signed-off-by: Alexander Bulekov <alxndr@bu.edu> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Darren Kenny <darren.kenny@oracle.com> Message-id: 20200220041118.23264-13-alxndr@bu.edu Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> --- MAINTAINERS | 8 ++ tests/qtest/fuzz/Makefile.include | 6 + tests/qtest/fuzz/fuzz.c | 179 ++++++++++++++++++++++++++++++ tests/qtest/fuzz/fuzz.h | 95 ++++++++++++++++ 4 files changed, 288 insertions(+) create mode 100644 tests/qtest/fuzz/Makefile.include create mode 100644 tests/qtest/fuzz/fuzz.c create mode 100644 tests/qtest/fuzz/fuzz.h diff --git a/MAINTAINERS b/MAINTAINERS index XXXXXXX..XXXXXXX 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -XXX,XX +XXX,XX @@ F: qtest.c F: accel/qtest.c F: tests/qtest/ +Device Fuzzing +M: Alexander Bulekov <alxndr@bu.edu> +R: Paolo Bonzini <pbonzini@redhat.com> +R: Bandan Das <bsd@redhat.com> +R: Stefan Hajnoczi <stefanha@redhat.com> +S: Maintained +F: tests/qtest/fuzz/ + Register API M: Alistair Francis <alistair@alistair23.me> S: Maintained diff --git a/tests/qtest/fuzz/Makefile.include b/tests/qtest/fuzz/Makefile.include new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/tests/qtest/fuzz/Makefile.include @@ -XXX,XX +XXX,XX @@ +QEMU_PROG_FUZZ=qemu-fuzz-$(TARGET_NAME)$(EXESUF) + +fuzz-obj-y += tests/qtest/libqtest.o +fuzz-obj-y += tests/qtest/fuzz/fuzz.o # Fuzzer skeleton + +FUZZ_CFLAGS += -I$(SRC_PATH)/tests -I$(SRC_PATH)/tests/qtest diff --git a/tests/qtest/fuzz/fuzz.c b/tests/qtest/fuzz/fuzz.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/tests/qtest/fuzz/fuzz.c @@ -XXX,XX +XXX,XX @@ +/* + * fuzzing driver + * + * Copyright Red Hat Inc., 2019 + * + * Authors: + * Alexander Bulekov <alxndr@bu.edu> + * + * This work is licensed under the terms of the GNU GPL, version 2 or later. + * See the COPYING file in the top-level directory. + * + */ + +#include "qemu/osdep.h" + +#include <wordexp.h> + +#include "sysemu/qtest.h" +#include "sysemu/runstate.h" +#include "sysemu/sysemu.h" +#include "qemu/main-loop.h" +#include "tests/qtest/libqtest.h" +#include "tests/qtest/libqos/qgraph.h" +#include "fuzz.h" + +#define MAX_EVENT_LOOPS 10 + +typedef struct FuzzTargetState { + FuzzTarget *target; + QSLIST_ENTRY(FuzzTargetState) target_list; +} FuzzTargetState; + +typedef QSLIST_HEAD(, FuzzTargetState) FuzzTargetList; + +static const char *fuzz_arch = TARGET_NAME; + +static FuzzTargetList *fuzz_target_list; +static FuzzTarget *fuzz_target; +static QTestState *fuzz_qts; + + + +void flush_events(QTestState *s) +{ + int i = MAX_EVENT_LOOPS; + while (g_main_context_pending(NULL) && i-- > 0) { + main_loop_wait(false); + } +} + +static QTestState *qtest_setup(void) +{ + qtest_server_set_send_handler(&qtest_client_inproc_recv, &fuzz_qts); + return qtest_inproc_init(&fuzz_qts, false, fuzz_arch, + &qtest_server_inproc_recv); +} + +void fuzz_add_target(const FuzzTarget *target) +{ + FuzzTargetState *tmp; + FuzzTargetState *target_state; + if (!fuzz_target_list) { + fuzz_target_list = g_new0(FuzzTargetList, 1); + } + + QSLIST_FOREACH(tmp, fuzz_target_list, target_list) { + if (g_strcmp0(tmp->target->name, target->name) == 0) { + fprintf(stderr, "Error: Fuzz target name %s already in use\n", + target->name); + abort(); + } + } + target_state = g_new0(FuzzTargetState, 1); + target_state->target = g_new0(FuzzTarget, 1); + *(target_state->target) = *target; + QSLIST_INSERT_HEAD(fuzz_target_list, target_state, target_list); +} + + + +static void usage(char *path) +{ + printf("Usage: %s --fuzz-target=FUZZ_TARGET [LIBFUZZER ARGUMENTS]\n", path); + printf("where FUZZ_TARGET is one of:\n"); + FuzzTargetState *tmp; + if (!fuzz_target_list) { + fprintf(stderr, "Fuzz target list not initialized\n"); + abort(); + } + QSLIST_FOREACH(tmp, fuzz_target_list, target_list) { + printf(" * %s : %s\n", tmp->target->name, + tmp->target->description); + } + exit(0); +} + +static FuzzTarget *fuzz_get_target(char* name) +{ + FuzzTargetState *tmp; + if (!fuzz_target_list) { + fprintf(stderr, "Fuzz target list not initialized\n"); + abort(); + } + + QSLIST_FOREACH(tmp, fuzz_target_list, target_list) { + if (strcmp(tmp->target->name, name) == 0) { + return tmp->target; + } + } + return NULL; +} + + +/* Executed for each fuzzing-input */ +int LLVMFuzzerTestOneInput(const unsigned char *Data, size_t Size) +{ + /* + * Do the pre-fuzz-initialization before the first fuzzing iteration, + * instead of before the actual fuzz loop. This is needed since libfuzzer + * may fork off additional workers, prior to the fuzzing loop, and if + * pre_fuzz() sets up e.g. shared memory, this should be done for the + * individual worker processes + */ + static int pre_fuzz_done; + if (!pre_fuzz_done && fuzz_target->pre_fuzz) { + fuzz_target->pre_fuzz(fuzz_qts); + pre_fuzz_done = true; + } + + fuzz_target->fuzz(fuzz_qts, Data, Size); + return 0; +} + +/* Executed once, prior to fuzzing */ +int LLVMFuzzerInitialize(int *argc, char ***argv, char ***envp) +{ + + char *target_name; + + /* Initialize qgraph and modules */ + qos_graph_init(); + module_call_init(MODULE_INIT_FUZZ_TARGET); + module_call_init(MODULE_INIT_QOM); + module_call_init(MODULE_INIT_LIBQOS); + + if (*argc <= 1) { + usage(**argv); + } + + /* Identify the fuzz target */ + target_name = (*argv)[1]; + if (!strstr(target_name, "--fuzz-target=")) { + usage(**argv); + } + + target_name += strlen("--fuzz-target="); + + fuzz_target = fuzz_get_target(target_name); + if (!fuzz_target) { + usage(**argv); + } + + fuzz_qts = qtest_setup(); + + if (fuzz_target->pre_vm_init) { + fuzz_target->pre_vm_init(); + } + + /* Run QEMU's softmmu main with the fuzz-target dependent arguments */ + const char *init_cmdline = fuzz_target->get_init_cmdline(fuzz_target); + + /* Split the runcmd into an argv and argc */ + wordexp_t result; + wordexp(init_cmdline, &result, 0); + + qemu_init(result.we_wordc, result.we_wordv, NULL); + + return 0; +} diff --git a/tests/qtest/fuzz/fuzz.h b/tests/qtest/fuzz/fuzz.h new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/tests/qtest/fuzz/fuzz.h @@ -XXX,XX +XXX,XX @@ +/* + * fuzzing driver + * + * Copyright Red Hat Inc., 2019 + * + * Authors: + * Alexander Bulekov <alxndr@bu.edu> + * + * This work is licensed under the terms of the GNU GPL, version 2 or later. + * See the COPYING file in the top-level directory. + * + */ + +#ifndef FUZZER_H_ +#define FUZZER_H_ + +#include "qemu/osdep.h" +#include "qemu/units.h" +#include "qapi/error.h" + +#include "tests/qtest/libqtest.h" + +/** + * A libfuzzer fuzzing target + * + * The QEMU fuzzing binary is built with all available targets, each + * with a unique @name that can be specified on the command-line to + * select which target should run. + * + * A target must implement ->fuzz() to process a random input. If QEMU + * crashes in ->fuzz() then libfuzzer will record a failure. + * + * Fuzzing targets are registered with fuzz_add_target(): + * + * static const FuzzTarget fuzz_target = { + * .name = "my-device-fifo", + * .description = "Fuzz the FIFO buffer registers of my-device", + * ... + * }; + * + * static void register_fuzz_target(void) + * { + * fuzz_add_target(&fuzz_target); + * } + * fuzz_target_init(register_fuzz_target); + */ +typedef struct FuzzTarget { + const char *name; /* target identifier (passed to --fuzz-target=)*/ + const char *description; /* help text */ + + + /* + * returns the arg-list that is passed to qemu/softmmu init() + * Cannot be NULL + */ + const char* (*get_init_cmdline)(struct FuzzTarget *); + + /* + * will run once, prior to running qemu/softmmu init. + * eg: set up shared-memory for communication with the child-process + * Can be NULL + */ + void(*pre_vm_init)(void); + + /* + * will run once, after QEMU has been initialized, prior to the fuzz-loop. + * eg: detect the memory map + * Can be NULL + */ + void(*pre_fuzz)(QTestState *); + + /* + * accepts and executes an input from libfuzzer. this is repeatedly + * executed during the fuzzing loop. Its should handle setup, input + * execution and cleanup. + * Cannot be NULL + */ + void(*fuzz)(QTestState *, const unsigned char *, size_t); + +} FuzzTarget; + +void flush_events(QTestState *); +void reboot(QTestState *); + +/* + * makes a copy of *target and adds it to the target-list. + * i.e. fine to set up target on the caller's stack + */ +void fuzz_add_target(const FuzzTarget *target); + +int LLVMFuzzerTestOneInput(const unsigned char *Data, size_t Size); +int LLVMFuzzerInitialize(int *argc, char ***argv, char ***envp); + +#endif + -- 2.24.1 From: Alexander Bulekov <alxndr@bu.edu> Ram blocks were marked MADV_DONTFORK breaking fuzzing-tests which execute each test-input in a forked process. Signed-off-by: Alexander Bulekov <alxndr@bu.edu> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Darren Kenny <darren.kenny@oracle.com> Message-id: 20200220041118.23264-14-alxndr@bu.edu Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> --- exec.c | 12 ++++++++++-- 1 file changed, 10 insertions(+), 2 deletions(-) diff --git a/exec.c b/exec.c index XXXXXXX..XXXXXXX 100644 --- a/exec.c +++ b/exec.c @@ -XXX,XX +XXX,XX @@ #include "sysemu/kvm.h" #include "sysemu/sysemu.h" #include "sysemu/tcg.h" +#include "sysemu/qtest.h" #include "qemu/timer.h" #include "qemu/config-file.h" #include "qemu/error-report.h" @@ -XXX,XX +XXX,XX @@ static void ram_block_add(RAMBlock *new_block, Error **errp, bool shared) if (new_block->host) { qemu_ram_setup_dump(new_block->host, new_block->max_length); qemu_madvise(new_block->host, new_block->max_length, QEMU_MADV_HUGEPAGE); - /* MADV_DONTFORK is also needed by KVM in absence of synchronous MMU */ - qemu_madvise(new_block->host, new_block->max_length, QEMU_MADV_DONTFORK); + /* + * MADV_DONTFORK is also needed by KVM in absence of synchronous MMU + * Configure it unless the machine is a qtest server, in which case + * KVM is not used and it may be forked (eg for fuzzing purposes). + */ + if (!qtest_enabled()) { + qemu_madvise(new_block->host, new_block->max_length, + QEMU_MADV_DONTFORK); + } ram_block_notify_add(new_block->host, new_block->max_length); } } -- 2.24.1 From: Alexander Bulekov <alxndr@bu.edu> The qtest-based fuzzer makes use of forking to reset-state between tests. Keep the callback enabled, so the call_rcu thread gets created within the child process. Signed-off-by: Alexander Bulekov <alxndr@bu.edu> Reviewed-by: Darren Kenny <darren.kenny@oracle.com> Acked-by: Stefan Hajnoczi <stefanha@redhat.com> Message-id: 20200220041118.23264-15-alxndr@bu.edu Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> --- softmmu/vl.c | 12 +++++++++++- 1 file changed, 11 insertions(+), 1 deletion(-) diff --git a/softmmu/vl.c b/softmmu/vl.c index XXXXXXX..XXXXXXX 100644 --- a/softmmu/vl.c +++ b/softmmu/vl.c @@ -XXX,XX +XXX,XX @@ void qemu_init(int argc, char **argv, char **envp) set_memory_options(&ram_slots, &maxram_size, machine_class); os_daemonize(); - rcu_disable_atfork(); + + /* + * If QTest is enabled, keep the rcu_atfork enabled, since system processes + * may be forked testing purposes (e.g. fork-server based fuzzing) The fork + * should happen before a signle cpu instruction is executed, to prevent + * deadlocks. See commit 73c6e40, rcu: "completely disable pthread_atfork + * callbacks as soon as possible" + */ + if (!qtest_enabled()) { + rcu_disable_atfork(); + } if (pid_file && !qemu_write_pidfile(pid_file, &err)) { error_reportf_err(err, "cannot create PID file: "); -- 2.24.1 From: Alexander Bulekov <alxndr@bu.edu> fork() is a simple way to ensure that state does not leak in between fuzzing runs. Unfortunately, the fuzzer mutation engine relies on bitmaps which contain coverage information for each fuzzing run, and these bitmaps should be copied from the child to the parent(where the mutation occurs). These bitmaps are created through compile-time instrumentation and they are not shared with fork()-ed processes, by default. To address this, we create a shared memory region, adjust its size and map it _over_ the counter region. Furthermore, libfuzzer doesn't generally expose the globals that specify the location of the counters/coverage bitmap. As a workaround, we rely on a custom linker script which forces all of the bitmaps we care about to be placed in a contiguous region, which is easy to locate and mmap over. Signed-off-by: Alexander Bulekov <alxndr@bu.edu> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Darren Kenny <darren.kenny@oracle.com> Message-id: 20200220041118.23264-16-alxndr@bu.edu Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> --- tests/qtest/fuzz/Makefile.include | 5 +++ tests/qtest/fuzz/fork_fuzz.c | 55 +++++++++++++++++++++++++++++++ tests/qtest/fuzz/fork_fuzz.h | 23 +++++++++++++ tests/qtest/fuzz/fork_fuzz.ld | 37 +++++++++++++++++++++ 4 files changed, 120 insertions(+) create mode 100644 tests/qtest/fuzz/fork_fuzz.c create mode 100644 tests/qtest/fuzz/fork_fuzz.h create mode 100644 tests/qtest/fuzz/fork_fuzz.ld diff --git a/tests/qtest/fuzz/Makefile.include b/tests/qtest/fuzz/Makefile.include index XXXXXXX..XXXXXXX 100644 --- a/tests/qtest/fuzz/Makefile.include +++ b/tests/qtest/fuzz/Makefile.include @@ -XXX,XX +XXX,XX @@ QEMU_PROG_FUZZ=qemu-fuzz-$(TARGET_NAME)$(EXESUF) fuzz-obj-y += tests/qtest/libqtest.o fuzz-obj-y += tests/qtest/fuzz/fuzz.o # Fuzzer skeleton +fuzz-obj-y += tests/qtest/fuzz/fork_fuzz.o FUZZ_CFLAGS += -I$(SRC_PATH)/tests -I$(SRC_PATH)/tests/qtest + +# Linker Script to force coverage-counters into known regions which we can mark +# shared +FUZZ_LDFLAGS += -Xlinker -T$(SRC_PATH)/tests/qtest/fuzz/fork_fuzz.ld diff --git a/tests/qtest/fuzz/fork_fuzz.c b/tests/qtest/fuzz/fork_fuzz.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/tests/qtest/fuzz/fork_fuzz.c @@ -XXX,XX +XXX,XX @@ +/* + * Fork-based fuzzing helpers + * + * Copyright Red Hat Inc., 2019 + * + * Authors: + * Alexander Bulekov <alxndr@bu.edu> + * + * This work is licensed under the terms of the GNU GPL, version 2 or later. + * See the COPYING file in the top-level directory. + * + */ + +#include "qemu/osdep.h" +#include "fork_fuzz.h" + + +void counter_shm_init(void) +{ + char *shm_path = g_strdup_printf("/qemu-fuzz-cntrs.%d", getpid()); + int fd = shm_open(shm_path, O_CREAT | O_RDWR, S_IRUSR | S_IWUSR); + g_free(shm_path); + + if (fd == -1) { + perror("Error: "); + exit(1); + } + if (ftruncate(fd, &__FUZZ_COUNTERS_END - &__FUZZ_COUNTERS_START) == -1) { + perror("Error: "); + exit(1); + } + /* Copy what's in the counter region to the shm.. */ + void *rptr = mmap(NULL , + &__FUZZ_COUNTERS_END - &__FUZZ_COUNTERS_START, + PROT_READ | PROT_WRITE, MAP_SHARED, fd, 0); + memcpy(rptr, + &__FUZZ_COUNTERS_START, + &__FUZZ_COUNTERS_END - &__FUZZ_COUNTERS_START); + + munmap(rptr, &__FUZZ_COUNTERS_END - &__FUZZ_COUNTERS_START); + + /* And map the shm over the counter region */ + rptr = mmap(&__FUZZ_COUNTERS_START, + &__FUZZ_COUNTERS_END - &__FUZZ_COUNTERS_START, + PROT_READ | PROT_WRITE, MAP_SHARED | MAP_FIXED, fd, 0); + + close(fd); + + if (!rptr) { + perror("Error: "); + exit(1); + } +} + + diff --git a/tests/qtest/fuzz/fork_fuzz.h b/tests/qtest/fuzz/fork_fuzz.h new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/tests/qtest/fuzz/fork_fuzz.h @@ -XXX,XX +XXX,XX @@ +/* + * Fork-based fuzzing helpers + * + * Copyright Red Hat Inc., 2019 + * + * Authors: + * Alexander Bulekov <alxndr@bu.edu> + * + * This work is licensed under the terms of the GNU GPL, version 2 or later. + * See the COPYING file in the top-level directory. + * + */ + +#ifndef FORK_FUZZ_H +#define FORK_FUZZ_H + +extern uint8_t __FUZZ_COUNTERS_START; +extern uint8_t __FUZZ_COUNTERS_END; + +void counter_shm_init(void); + +#endif + diff --git a/tests/qtest/fuzz/fork_fuzz.ld b/tests/qtest/fuzz/fork_fuzz.ld new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/tests/qtest/fuzz/fork_fuzz.ld @@ -XXX,XX +XXX,XX @@ +/* We adjust linker script modification to place all of the stuff that needs to + * persist across fuzzing runs into a contiguous seciton of memory. Then, it is + * easy to re-map the counter-related memory as shared. +*/ + +SECTIONS +{ + .data.fuzz_start : ALIGN(4K) + { + __FUZZ_COUNTERS_START = .; + __start___sancov_cntrs = .; + *(_*sancov_cntrs); + __stop___sancov_cntrs = .; + + /* Lowest stack counter */ + *(__sancov_lowest_stack); + } + .data.fuzz_ordered : + { + /* Coverage counters. They're not necessary for fuzzing, but are useful + * for analyzing the fuzzing performance + */ + __start___llvm_prf_cnts = .; + *(*llvm_prf_cnts); + __stop___llvm_prf_cnts = .; + + /* Internal Libfuzzer TracePC object which contains the ValueProfileMap */ + FuzzerTracePC*(.bss*); + } + .data.fuzz_end : ALIGN(4K) + { + __FUZZ_COUNTERS_END = .; + } +} +/* Dont overwrite the SECTIONS in the default linker script. Instead insert the + * above into the default script */ +INSERT AFTER .data; -- 2.24.1 From: Alexander Bulekov <alxndr@bu.edu> Signed-off-by: Alexander Bulekov <alxndr@bu.edu> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Darren Kenny <darren.kenny@oracle.com> Message-id: 20200220041118.23264-17-alxndr@bu.edu Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> --- tests/qtest/fuzz/Makefile.include | 2 + tests/qtest/fuzz/qos_fuzz.c | 234 ++++++++++++++++++++++++++++++ tests/qtest/fuzz/qos_fuzz.h | 33 +++++ 3 files changed, 269 insertions(+) create mode 100644 tests/qtest/fuzz/qos_fuzz.c create mode 100644 tests/qtest/fuzz/qos_fuzz.h diff --git a/tests/qtest/fuzz/Makefile.include b/tests/qtest/fuzz/Makefile.include index XXXXXXX..XXXXXXX 100644 --- a/tests/qtest/fuzz/Makefile.include +++ b/tests/qtest/fuzz/Makefile.include @@ -XXX,XX +XXX,XX @@ QEMU_PROG_FUZZ=qemu-fuzz-$(TARGET_NAME)$(EXESUF) fuzz-obj-y += tests/qtest/libqtest.o +fuzz-obj-y += $(libqos-obj-y) fuzz-obj-y += tests/qtest/fuzz/fuzz.o # Fuzzer skeleton fuzz-obj-y += tests/qtest/fuzz/fork_fuzz.o +fuzz-obj-y += tests/qtest/fuzz/qos_fuzz.o FUZZ_CFLAGS += -I$(SRC_PATH)/tests -I$(SRC_PATH)/tests/qtest diff --git a/tests/qtest/fuzz/qos_fuzz.c b/tests/qtest/fuzz/qos_fuzz.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/tests/qtest/fuzz/qos_fuzz.c @@ -XXX,XX +XXX,XX @@ +/* + * QOS-assisted fuzzing helpers + * + * Copyright (c) 2018 Emanuele Giuseppe Esposito <e.emanuelegiuseppe@gmail.com> + * + * This library is free software; you can redistribute it and/or + * modify it under the terms of the GNU Lesser General Public + * License version 2 as published by the Free Software Foundation. + * + * This library is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public + * License along with this library; if not, see <http://www.gnu.org/licenses/> + */ + +#include "qemu/osdep.h" +#include "qemu/units.h" +#include "qapi/error.h" +#include "qemu-common.h" +#include "exec/memory.h" +#include "exec/address-spaces.h" +#include "sysemu/sysemu.h" +#include "qemu/main-loop.h" + +#include "tests/qtest/libqtest.h" +#include "tests/qtest/libqos/malloc.h" +#include "tests/qtest/libqos/qgraph.h" +#include "tests/qtest/libqos/qgraph_internal.h" +#include "tests/qtest/libqos/qos_external.h" + +#include "fuzz.h" +#include "qos_fuzz.h" + +#include "qapi/qapi-commands-machine.h" +#include "qapi/qapi-commands-qom.h" +#include "qapi/qmp/qlist.h" + + +void *fuzz_qos_obj; +QGuestAllocator *fuzz_qos_alloc; + +static const char *fuzz_target_name; +static char **fuzz_path_vec; + +/* + * Replaced the qmp commands with direct qmp_marshal calls. + * Probably there is a better way to do this + */ +static void qos_set_machines_devices_available(void) +{ + QDict *req = qdict_new(); + QObject *response; + QDict *args = qdict_new(); + QList *lst; + Error *err = NULL; + + qmp_marshal_query_machines(NULL, &response, &err); + assert(!err); + lst = qobject_to(QList, response); + apply_to_qlist(lst, true); + + qobject_unref(response); + + + qdict_put_str(req, "execute", "qom-list-types"); + qdict_put_str(args, "implements", "device"); + qdict_put_bool(args, "abstract", true); + qdict_put_obj(req, "arguments", (QObject *) args); + + qmp_marshal_qom_list_types(args, &response, &err); + assert(!err); + lst = qobject_to(QList, response); + apply_to_qlist(lst, false); + qobject_unref(response); + qobject_unref(req); +} + +static char **current_path; + +void *qos_allocate_objects(QTestState *qts, QGuestAllocator **p_alloc) +{ + return allocate_objects(qts, current_path + 1, p_alloc); +} + +static const char *qos_build_main_args(void) +{ + char **path = fuzz_path_vec; + QOSGraphNode *test_node; + GString *cmd_line = g_string_new(path[0]); + void *test_arg; + + if (!path) { + fprintf(stderr, "QOS Path not found\n"); + abort(); + } + + /* Before test */ + current_path = path; + test_node = qos_graph_get_node(path[(g_strv_length(path) - 1)]); + test_arg = test_node->u.test.arg; + if (test_node->u.test.before) { + test_arg = test_node->u.test.before(cmd_line, test_arg); + } + /* Prepend the arguments that we need */ + g_string_prepend(cmd_line, + TARGET_NAME " -display none -machine accel=qtest -m 64 "); + return cmd_line->str; +} + +/* + * This function is largely a copy of qos-test.c:walk_path. Since walk_path + * is itself a callback, its a little annoying to add another argument/layer of + * indirection + */ +static void walk_path(QOSGraphNode *orig_path, int len) +{ + QOSGraphNode *path; + QOSGraphEdge *edge; + + /* etype set to QEDGE_CONSUMED_BY so that machine can add to the command line */ + QOSEdgeType etype = QEDGE_CONSUMED_BY; + + /* twice QOS_PATH_MAX_ELEMENT_SIZE since each edge can have its arg */ + char **path_vec = g_new0(char *, (QOS_PATH_MAX_ELEMENT_SIZE * 2)); + int path_vec_size = 0; + + char *after_cmd, *before_cmd, *after_device; + GString *after_device_str = g_string_new(""); + char *node_name = orig_path->name, *path_str; + + GString *cmd_line = g_string_new(""); + GString *cmd_line2 = g_string_new(""); + + path = qos_graph_get_node(node_name); /* root */ + node_name = qos_graph_edge_get_dest(path->path_edge); /* machine name */ + + path_vec[path_vec_size++] = node_name; + path_vec[path_vec_size++] = qos_get_machine_type(node_name); + + for (;;) { + path = qos_graph_get_node(node_name); + if (!path->path_edge) { + break; + } + + node_name = qos_graph_edge_get_dest(path->path_edge); + + /* append node command line + previous edge command line */ + if (path->command_line && etype == QEDGE_CONSUMED_BY) { + g_string_append(cmd_line, path->command_line); + g_string_append(cmd_line, after_device_str->str); + g_string_truncate(after_device_str, 0); + } + + path_vec[path_vec_size++] = qos_graph_edge_get_name(path->path_edge); + /* detect if edge has command line args */ + after_cmd = qos_graph_edge_get_after_cmd_line(path->path_edge); + after_device = qos_graph_edge_get_extra_device_opts(path->path_edge); + before_cmd = qos_graph_edge_get_before_cmd_line(path->path_edge); + edge = qos_graph_get_edge(path->name, node_name); + etype = qos_graph_edge_get_type(edge); + + if (before_cmd) { + g_string_append(cmd_line, before_cmd); + } + if (after_cmd) { + g_string_append(cmd_line2, after_cmd); + } + if (after_device) { + g_string_append(after_device_str, after_device); + } + } + + path_vec[path_vec_size++] = NULL; + g_string_append(cmd_line, after_device_str->str); + g_string_free(after_device_str, true); + + g_string_append(cmd_line, cmd_line2->str); + g_string_free(cmd_line2, true); + + /* + * here position 0 has <arch>/<machine>, position 1 has <machine>. + * The path must not have the <arch>, qtest_add_data_func adds it. + */ + path_str = g_strjoinv("/", path_vec + 1); + + /* Check that this is the test we care about: */ + char *test_name = strrchr(path_str, '/') + 1; + if (strcmp(test_name, fuzz_target_name) == 0) { + /* + * put arch/machine in position 1 so run_one_test can do its work + * and add the command line at position 0. + */ + path_vec[1] = path_vec[0]; + path_vec[0] = g_string_free(cmd_line, false); + + fuzz_path_vec = path_vec; + } else { + g_free(path_vec); + } + + g_free(path_str); +} + +static const char *qos_get_cmdline(FuzzTarget *t) +{ + /* + * Set a global variable that we use to identify the qos_path for our + * fuzz_target + */ + fuzz_target_name = t->name; + qos_set_machines_devices_available(); + qos_graph_foreach_test_path(walk_path); + return qos_build_main_args(); +} + +void fuzz_add_qos_target( + FuzzTarget *fuzz_opts, + const char *interface, + QOSGraphTestOptions *opts + ) +{ + qos_add_test(fuzz_opts->name, interface, NULL, opts); + fuzz_opts->get_init_cmdline = qos_get_cmdline; + fuzz_add_target(fuzz_opts); +} + +void qos_init_path(QTestState *s) +{ + fuzz_qos_obj = qos_allocate_objects(s , &fuzz_qos_alloc); +} diff --git a/tests/qtest/fuzz/qos_fuzz.h b/tests/qtest/fuzz/qos_fuzz.h new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/tests/qtest/fuzz/qos_fuzz.h @@ -XXX,XX +XXX,XX @@ +/* + * QOS-assisted fuzzing helpers + * + * Copyright Red Hat Inc., 2019 + * + * Authors: + * Alexander Bulekov <alxndr@bu.edu> + * + * This work is licensed under the terms of the GNU GPL, version 2 or later. + * See the COPYING file in the top-level directory. + */ + +#ifndef _QOS_FUZZ_H_ +#define _QOS_FUZZ_H_ + +#include "tests/qtest/fuzz/fuzz.h" +#include "tests/qtest/libqos/qgraph.h" + +int qos_fuzz(const unsigned char *Data, size_t Size); +void qos_setup(void); + +extern void *fuzz_qos_obj; +extern QGuestAllocator *fuzz_qos_alloc; + +void fuzz_add_qos_target( + FuzzTarget *fuzz_opts, + const char *interface, + QOSGraphTestOptions *opts + ); + +void qos_init_path(QTestState *); + +#endif -- 2.24.1 From: Alexander Bulekov <alxndr@bu.edu> Signed-off-by: Alexander Bulekov <alxndr@bu.edu> Reviewed-by: Darren Kenny <darren.kenny@oracle.com> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Message-id: 20200220041118.23264-18-alxndr@bu.edu Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> --- Makefile | 15 ++++++++++++++- Makefile.target | 16 ++++++++++++++++ 2 files changed, 30 insertions(+), 1 deletion(-) diff --git a/Makefile b/Makefile index XXXXXXX..XXXXXXX 100644 --- a/Makefile +++ b/Makefile @@ -XXX,XX +XXX,XX @@ config-host.h-timestamp: config-host.mak qemu-options.def: $(SRC_PATH)/qemu-options.hx $(SRC_PATH)/scripts/hxtool $(call quiet-command,sh $(SRC_PATH)/scripts/hxtool -h < $< > $@,"GEN","$@") -TARGET_DIRS_RULES := $(foreach t, all clean install, $(addsuffix /$(t), $(TARGET_DIRS))) +TARGET_DIRS_RULES := $(foreach t, all fuzz clean install, $(addsuffix /$(t), $(TARGET_DIRS))) SOFTMMU_ALL_RULES=$(filter %-softmmu/all, $(TARGET_DIRS_RULES)) $(SOFTMMU_ALL_RULES): $(authz-obj-y) @@ -XXX,XX +XXX,XX @@ ifdef DECOMPRESS_EDK2_BLOBS $(SOFTMMU_ALL_RULES): $(edk2-decompressed) endif +SOFTMMU_FUZZ_RULES=$(filter %-softmmu/fuzz, $(TARGET_DIRS_RULES)) +$(SOFTMMU_FUZZ_RULES): $(authz-obj-y) +$(SOFTMMU_FUZZ_RULES): $(block-obj-y) +$(SOFTMMU_FUZZ_RULES): $(chardev-obj-y) +$(SOFTMMU_FUZZ_RULES): $(crypto-obj-y) +$(SOFTMMU_FUZZ_RULES): $(io-obj-y) +$(SOFTMMU_FUZZ_RULES): config-all-devices.mak +$(SOFTMMU_FUZZ_RULES): $(edk2-decompressed) + .PHONY: $(TARGET_DIRS_RULES) # The $(TARGET_DIRS_RULES) are of the form SUBDIR/GOAL, so that # $(dir $@) yields the sub-directory, and $(notdir $@) yields the sub-goal @@ -XXX,XX +XXX,XX @@ subdir-slirp: slirp/all $(filter %/all, $(TARGET_DIRS_RULES)): libqemuutil.a $(common-obj-y) \ $(qom-obj-y) +$(filter %/fuzz, $(TARGET_DIRS_RULES)): libqemuutil.a $(common-obj-y) \ + $(qom-obj-y) $(crypto-user-obj-$(CONFIG_USER_ONLY)) + ROM_DIRS = $(addprefix pc-bios/, $(ROMS)) ROM_DIRS_RULES=$(foreach t, all clean, $(addsuffix /$(t), $(ROM_DIRS))) # Only keep -O and -g cflags @@ -XXX,XX +XXX,XX @@ $(ROM_DIRS_RULES): .PHONY: recurse-all recurse-clean recurse-install recurse-all: $(addsuffix /all, $(TARGET_DIRS) $(ROM_DIRS)) +recurse-fuzz: $(addsuffix /fuzz, $(TARGET_DIRS) $(ROM_DIRS)) recurse-clean: $(addsuffix /clean, $(TARGET_DIRS) $(ROM_DIRS)) recurse-install: $(addsuffix /install, $(TARGET_DIRS)) $(addsuffix /install, $(TARGET_DIRS)): all diff --git a/Makefile.target b/Makefile.target index XXXXXXX..XXXXXXX 100644 --- a/Makefile.target +++ b/Makefile.target @@ -XXX,XX +XXX,XX @@ ifdef CONFIG_TRACE_SYSTEMTAP rm -f *.stp endif +ifdef CONFIG_FUZZ +include $(SRC_PATH)/tests/qtest/fuzz/Makefile.include +include $(SRC_PATH)/tests/qtest/Makefile.include + +fuzz: fuzz-vars +fuzz-vars: QEMU_CFLAGS := $(FUZZ_CFLAGS) $(QEMU_CFLAGS) +fuzz-vars: QEMU_LDFLAGS := $(FUZZ_LDFLAGS) $(QEMU_LDFLAGS) +fuzz-vars: $(QEMU_PROG_FUZZ) +dummy := $(call unnest-vars,, fuzz-obj-y) + + +$(QEMU_PROG_FUZZ): config-devices.mak $(all-obj-y) $(COMMON_LDADDS) $(fuzz-obj-y) + $(call LINK, $(filter-out %.mak, $^)) + +endif + install: all ifneq ($(PROGS),) $(call install-prog,$(PROGS),$(DESTDIR)$(bindir)) -- 2.24.1 From: Alexander Bulekov <alxndr@bu.edu> Signed-off-by: Alexander Bulekov <alxndr@bu.edu> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com> Reviewed-by: Darren Kenny <darren.kenny@oracle.com> Message-id: 20200220041118.23264-19-alxndr@bu.edu Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> --- configure | 39 +++++++++++++++++++++++++++++++++++++++ 1 file changed, 39 insertions(+) diff --git a/configure b/configure index XXXXXXX..XXXXXXX 100755 --- a/configure +++ b/configure @@ -XXX,XX +XXX,XX @@ debug_mutex="no" libpmem="" default_devices="yes" plugins="no" +fuzzing="no" supported_cpu="no" supported_os="no" @@ -XXX,XX +XXX,XX @@ int main(void) { return 0; } EOF } +write_c_fuzzer_skeleton() { + cat > $TMPC <<EOF +#include <stdint.h> +#include <sys/types.h> +int LLVMFuzzerTestOneInput(const uint8_t *Data, size_t Size); +int LLVMFuzzerTestOneInput(const uint8_t *Data, size_t Size) { return 0; } +EOF +} + if check_define __linux__ ; then targetos="Linux" elif check_define _WIN32 ; then @@ -XXX,XX +XXX,XX @@ for opt do ;; --disable-containers) use_containers="no" ;; + --enable-fuzzing) fuzzing=yes + ;; + --disable-fuzzing) fuzzing=no + ;; *) echo "ERROR: unknown option $opt" echo "Try '$0 --help' for more information" @@ -XXX,XX +XXX,XX @@ EOF fi fi +########################################## +# checks for fuzzer +if test "$fuzzing" = "yes" ; then + write_c_fuzzer_skeleton + if compile_prog "$CPU_CFLAGS -Werror -fsanitize=address,fuzzer" ""; then + have_fuzzer=yes + fi +fi + ########################################## # check for libpmem @@ -XXX,XX +XXX,XX @@ echo "libpmem support $libpmem" echo "libudev $libudev" echo "default devices $default_devices" echo "plugin support $plugins" +echo "fuzzing support $fuzzing" if test "$supported_cpu" = "no"; then echo @@ -XXX,XX +XXX,XX @@ fi if test "$sheepdog" = "yes" ; then echo "CONFIG_SHEEPDOG=y" >> $config_host_mak fi +if test "$fuzzing" = "yes" ; then + if test "$have_fuzzer" = "yes"; then + FUZZ_LDFLAGS=" -fsanitize=address,fuzzer" + FUZZ_CFLAGS=" -fsanitize=address,fuzzer" + CFLAGS=" -fsanitize=address,fuzzer-no-link" + else + error_exit "Your compiler doesn't support -fsanitize=address,fuzzer" + exit 1 + fi +fi if test "$plugins" = "yes" ; then echo "CONFIG_PLUGIN=y" >> $config_host_mak @@ -XXX,XX +XXX,XX @@ if test "$libudev" != "no"; then echo "CONFIG_LIBUDEV=y" >> $config_host_mak echo "LIBUDEV_LIBS=$libudev_libs" >> $config_host_mak fi +if test "$fuzzing" != "no"; then + echo "CONFIG_FUZZ=y" >> $config_host_mak + echo "FUZZ_CFLAGS=$FUZZ_CFLAGS" >> $config_host_mak + echo "FUZZ_LDFLAGS=$FUZZ_LDFLAGS" >> $config_host_mak +fi if test "$edk2_blobs" = "yes" ; then echo "DECOMPRESS_EDK2_BLOBS=y" >> $config_host_mak -- 2.24.1 From: Alexander Bulekov <alxndr@bu.edu> These three targets should simply fuzz reads/writes to a couple ioports, but they mostly serve as examples of different ways to write targets. They demonstrate using qtest and qos for fuzzing, as well as using rebooting and forking to reset state, or not resetting it at all. Signed-off-by: Alexander Bulekov <alxndr@bu.edu> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Darren Kenny <darren.kenny@oracle.com> Message-id: 20200220041118.23264-20-alxndr@bu.edu Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> --- tests/qtest/fuzz/Makefile.include | 3 + tests/qtest/fuzz/i440fx_fuzz.c | 193 ++++++++++++++++++++++++++++++ 2 files changed, 196 insertions(+) create mode 100644 tests/qtest/fuzz/i440fx_fuzz.c diff --git a/tests/qtest/fuzz/Makefile.include b/tests/qtest/fuzz/Makefile.include index XXXXXXX..XXXXXXX 100644 --- a/tests/qtest/fuzz/Makefile.include +++ b/tests/qtest/fuzz/Makefile.include @@ -XXX,XX +XXX,XX @@ fuzz-obj-y += tests/qtest/fuzz/fuzz.o # Fuzzer skeleton fuzz-obj-y += tests/qtest/fuzz/fork_fuzz.o fuzz-obj-y += tests/qtest/fuzz/qos_fuzz.o +# Targets +fuzz-obj-y += tests/qtest/fuzz/i440fx_fuzz.o + FUZZ_CFLAGS += -I$(SRC_PATH)/tests -I$(SRC_PATH)/tests/qtest # Linker Script to force coverage-counters into known regions which we can mark diff --git a/tests/qtest/fuzz/i440fx_fuzz.c b/tests/qtest/fuzz/i440fx_fuzz.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/tests/qtest/fuzz/i440fx_fuzz.c @@ -XXX,XX +XXX,XX @@ +/* + * I440FX Fuzzing Target + * + * Copyright Red Hat Inc., 2019 + * + * Authors: + * Alexander Bulekov <alxndr@bu.edu> + * + * This work is licensed under the terms of the GNU GPL, version 2 or later. + * See the COPYING file in the top-level directory. + */ + +#include "qemu/osdep.h" + +#include "qemu/main-loop.h" +#include "tests/qtest/libqtest.h" +#include "tests/qtest/libqos/pci.h" +#include "tests/qtest/libqos/pci-pc.h" +#include "fuzz.h" +#include "fuzz/qos_fuzz.h" +#include "fuzz/fork_fuzz.h" + + +#define I440FX_PCI_HOST_BRIDGE_CFG 0xcf8 +#define I440FX_PCI_HOST_BRIDGE_DATA 0xcfc + +/* + * the input to the fuzzing functions below is a buffer of random bytes. we + * want to convert these bytes into a sequence of qtest or qos calls. to do + * this we define some opcodes: + */ +enum action_id { + WRITEB, + WRITEW, + WRITEL, + READB, + READW, + READL, + ACTION_MAX +}; + +static void i440fx_fuzz_qtest(QTestState *s, + const unsigned char *Data, size_t Size) { + /* + * loop over the Data, breaking it up into actions. each action has an + * opcode, address offset and value + */ + typedef struct QTestFuzzAction { + uint8_t opcode; + uint8_t addr; + uint32_t value; + } QTestFuzzAction; + QTestFuzzAction a; + + while (Size >= sizeof(a)) { + /* make a copy of the action so we can normalize the values in-place */ + memcpy(&a, Data, sizeof(a)); + /* select between two i440fx Port IO addresses */ + uint16_t addr = a.addr % 2 ? I440FX_PCI_HOST_BRIDGE_CFG : + I440FX_PCI_HOST_BRIDGE_DATA; + switch (a.opcode % ACTION_MAX) { + case WRITEB: + qtest_outb(s, addr, (uint8_t)a.value); + break; + case WRITEW: + qtest_outw(s, addr, (uint16_t)a.value); + break; + case WRITEL: + qtest_outl(s, addr, (uint32_t)a.value); + break; + case READB: + qtest_inb(s, addr); + break; + case READW: + qtest_inw(s, addr); + break; + case READL: + qtest_inl(s, addr); + break; + } + /* Move to the next operation */ + Size -= sizeof(a); + Data += sizeof(a); + } + flush_events(s); +} + +static void i440fx_fuzz_qos(QTestState *s, + const unsigned char *Data, size_t Size) { + /* + * Same as i440fx_fuzz_qtest, but using QOS. devfn is incorporated into the + * value written over Port IO + */ + typedef struct QOSFuzzAction { + uint8_t opcode; + uint8_t offset; + int devfn; + uint32_t value; + } QOSFuzzAction; + + static QPCIBus *bus; + if (!bus) { + bus = qpci_new_pc(s, fuzz_qos_alloc); + } + + QOSFuzzAction a; + while (Size >= sizeof(a)) { + memcpy(&a, Data, sizeof(a)); + switch (a.opcode % ACTION_MAX) { + case WRITEB: + bus->config_writeb(bus, a.devfn, a.offset, (uint8_t)a.value); + break; + case WRITEW: + bus->config_writew(bus, a.devfn, a.offset, (uint16_t)a.value); + break; + case WRITEL: + bus->config_writel(bus, a.devfn, a.offset, (uint32_t)a.value); + break; + case READB: + bus->config_readb(bus, a.devfn, a.offset); + break; + case READW: + bus->config_readw(bus, a.devfn, a.offset); + break; + case READL: + bus->config_readl(bus, a.devfn, a.offset); + break; + } + Size -= sizeof(a); + Data += sizeof(a); + } + flush_events(s); +} + +static void i440fx_fuzz_qos_fork(QTestState *s, + const unsigned char *Data, size_t Size) { + if (fork() == 0) { + i440fx_fuzz_qos(s, Data, Size); + _Exit(0); + } else { + wait(NULL); + } +} + +static const char *i440fx_qtest_argv = TARGET_NAME " -machine accel=qtest" + "-m 0 -display none"; +static const char *i440fx_argv(FuzzTarget *t) +{ + return i440fx_qtest_argv; +} + +static void fork_init(void) +{ + counter_shm_init(); +} + +static void register_pci_fuzz_targets(void) +{ + /* Uses simple qtest commands and reboots to reset state */ + fuzz_add_target(&(FuzzTarget){ + .name = "i440fx-qtest-reboot-fuzz", + .description = "Fuzz the i440fx using raw qtest commands and" + "rebooting after each run", + .get_init_cmdline = i440fx_argv, + .fuzz = i440fx_fuzz_qtest}); + + /* Uses libqos and forks to prevent state leakage */ + fuzz_add_qos_target(&(FuzzTarget){ + .name = "i440fx-qos-fork-fuzz", + .description = "Fuzz the i440fx using raw qtest commands and" + "rebooting after each run", + .pre_vm_init = &fork_init, + .fuzz = i440fx_fuzz_qos_fork,}, + "i440FX-pcihost", + &(QOSGraphTestOptions){} + ); + + /* + * Uses libqos. Doesn't do anything to reset state. Note that if we were to + * reboot after each run, we would also have to redo the qos-related + * initialization (qos_init_path) + */ + fuzz_add_qos_target(&(FuzzTarget){ + .name = "i440fx-qos-noreset-fuzz", + .description = "Fuzz the i440fx using raw qtest commands and" + "rebooting after each run", + .fuzz = i440fx_fuzz_qos,}, + "i440FX-pcihost", + &(QOSGraphTestOptions){} + ); +} + +fuzz_target_init(register_pci_fuzz_targets); -- 2.24.1 From: Alexander Bulekov <alxndr@bu.edu> The virtio-net fuzz target feeds inputs to all three virtio-net virtqueues, and uses forking to avoid leaking state between fuzz runs. Signed-off-by: Alexander Bulekov <alxndr@bu.edu> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Darren Kenny <darren.kenny@oracle.com> Message-id: 20200220041118.23264-21-alxndr@bu.edu Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> --- tests/qtest/fuzz/Makefile.include | 1 + tests/qtest/fuzz/virtio_net_fuzz.c | 198 +++++++++++++++++++++++++++++ 2 files changed, 199 insertions(+) create mode 100644 tests/qtest/fuzz/virtio_net_fuzz.c diff --git a/tests/qtest/fuzz/Makefile.include b/tests/qtest/fuzz/Makefile.include index XXXXXXX..XXXXXXX 100644 --- a/tests/qtest/fuzz/Makefile.include +++ b/tests/qtest/fuzz/Makefile.include @@ -XXX,XX +XXX,XX @@ fuzz-obj-y += tests/qtest/fuzz/qos_fuzz.o # Targets fuzz-obj-y += tests/qtest/fuzz/i440fx_fuzz.o +fuzz-obj-y += tests/qtest/fuzz/virtio_net_fuzz.o FUZZ_CFLAGS += -I$(SRC_PATH)/tests -I$(SRC_PATH)/tests/qtest diff --git a/tests/qtest/fuzz/virtio_net_fuzz.c b/tests/qtest/fuzz/virtio_net_fuzz.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/tests/qtest/fuzz/virtio_net_fuzz.c @@ -XXX,XX +XXX,XX @@ +/* + * virtio-net Fuzzing Target + * + * Copyright Red Hat Inc., 2019 + * + * Authors: + * Alexander Bulekov <alxndr@bu.edu> + * + * This work is licensed under the terms of the GNU GPL, version 2 or later. + * See the COPYING file in the top-level directory. + */ + +#include "qemu/osdep.h" + +#include "standard-headers/linux/virtio_config.h" +#include "tests/qtest/libqtest.h" +#include "tests/qtest/libqos/virtio-net.h" +#include "fuzz.h" +#include "fork_fuzz.h" +#include "qos_fuzz.h" + + +#define QVIRTIO_NET_TIMEOUT_US (30 * 1000 * 1000) +#define QVIRTIO_RX_VQ 0 +#define QVIRTIO_TX_VQ 1 +#define QVIRTIO_CTRL_VQ 2 + +static int sockfds[2]; +static bool sockfds_initialized; + +static void virtio_net_fuzz_multi(QTestState *s, + const unsigned char *Data, size_t Size, bool check_used) +{ + typedef struct vq_action { + uint8_t queue; + uint8_t length; + uint8_t write; + uint8_t next; + uint8_t rx; + } vq_action; + + uint32_t free_head = 0; + + QGuestAllocator *t_alloc = fuzz_qos_alloc; + + QVirtioNet *net_if = fuzz_qos_obj; + QVirtioDevice *dev = net_if->vdev; + QVirtQueue *q; + vq_action vqa; + while (Size >= sizeof(vqa)) { + memcpy(&vqa, Data, sizeof(vqa)); + Data += sizeof(vqa); + Size -= sizeof(vqa); + + q = net_if->queues[vqa.queue % 3]; + + vqa.length = vqa.length >= Size ? Size : vqa.length; + + /* + * Only attempt to write incoming packets, when using the socket + * backend. Otherwise, always place the input on a virtqueue. + */ + if (vqa.rx && sockfds_initialized) { + write(sockfds[0], Data, vqa.length); + } else { + vqa.rx = 0; + uint64_t req_addr = guest_alloc(t_alloc, vqa.length); + /* + * If checking used ring, ensure that the fuzzer doesn't trigger + * trivial asserion failure on zero-zied buffer + */ + qtest_memwrite(s, req_addr, Data, vqa.length); + + + free_head = qvirtqueue_add(s, q, req_addr, vqa.length, + vqa.write, vqa.next); + qvirtqueue_add(s, q, req_addr, vqa.length, vqa.write , vqa.next); + qvirtqueue_kick(s, dev, q, free_head); + } + + /* Run the main loop */ + qtest_clock_step(s, 100); + flush_events(s); + + /* Wait on used descriptors */ + if (check_used && !vqa.rx) { + gint64 start_time = g_get_monotonic_time(); + /* + * normally, we could just use qvirtio_wait_used_elem, but since we + * must manually run the main-loop for all the bhs to run, we use + * this hack with flush_events(), to run the main_loop + */ + while (!vqa.rx && q != net_if->queues[QVIRTIO_RX_VQ]) { + uint32_t got_desc_idx; + /* Input led to a virtio_error */ + if (dev->bus->get_status(dev) & VIRTIO_CONFIG_S_NEEDS_RESET) { + break; + } + if (dev->bus->get_queue_isr_status(dev, q) && + qvirtqueue_get_buf(s, q, &got_desc_idx, NULL)) { + g_assert_cmpint(got_desc_idx, ==, free_head); + break; + } + g_assert(g_get_monotonic_time() - start_time + <= QVIRTIO_NET_TIMEOUT_US); + + /* Run the main loop */ + qtest_clock_step(s, 100); + flush_events(s); + } + } + Data += vqa.length; + Size -= vqa.length; + } +} + +static void virtio_net_fork_fuzz(QTestState *s, + const unsigned char *Data, size_t Size) +{ + if (fork() == 0) { + virtio_net_fuzz_multi(s, Data, Size, false); + flush_events(s); + _Exit(0); + } else { + wait(NULL); + } +} + +static void virtio_net_fork_fuzz_check_used(QTestState *s, + const unsigned char *Data, size_t Size) +{ + if (fork() == 0) { + virtio_net_fuzz_multi(s, Data, Size, true); + flush_events(s); + _Exit(0); + } else { + wait(NULL); + } +} + +static void virtio_net_pre_fuzz(QTestState *s) +{ + qos_init_path(s); + counter_shm_init(); +} + +static void *virtio_net_test_setup_socket(GString *cmd_line, void *arg) +{ + int ret = socketpair(PF_UNIX, SOCK_STREAM, 0, sockfds); + g_assert_cmpint(ret, !=, -1); + fcntl(sockfds[0], F_SETFL, O_NONBLOCK); + sockfds_initialized = true; + g_string_append_printf(cmd_line, " -netdev socket,fd=%d,id=hs0 ", + sockfds[1]); + return arg; +} + +static void *virtio_net_test_setup_user(GString *cmd_line, void *arg) +{ + g_string_append_printf(cmd_line, " -netdev user,id=hs0 "); + return arg; +} + +static void register_virtio_net_fuzz_targets(void) +{ + fuzz_add_qos_target(&(FuzzTarget){ + .name = "virtio-net-socket", + .description = "Fuzz the virtio-net virtual queues. Fuzz incoming " + "traffic using the socket backend", + .pre_fuzz = &virtio_net_pre_fuzz, + .fuzz = virtio_net_fork_fuzz,}, + "virtio-net", + &(QOSGraphTestOptions){.before = virtio_net_test_setup_socket} + ); + + fuzz_add_qos_target(&(FuzzTarget){ + .name = "virtio-net-socket-check-used", + .description = "Fuzz the virtio-net virtual queues. Wait for the " + "descriptors to be used. Timeout may indicate improperly handled " + "input", + .pre_fuzz = &virtio_net_pre_fuzz, + .fuzz = virtio_net_fork_fuzz_check_used,}, + "virtio-net", + &(QOSGraphTestOptions){.before = virtio_net_test_setup_socket} + ); + fuzz_add_qos_target(&(FuzzTarget){ + .name = "virtio-net-slirp", + .description = "Fuzz the virtio-net virtual queues with the slirp " + " backend. Warning: May result in network traffic emitted from the " + " process. Run in an isolated network environment.", + .pre_fuzz = &virtio_net_pre_fuzz, + .fuzz = virtio_net_fork_fuzz,}, + "virtio-net", + &(QOSGraphTestOptions){.before = virtio_net_test_setup_user} + ); +} + +fuzz_target_init(register_virtio_net_fuzz_targets); -- 2.24.1 From: Alexander Bulekov <alxndr@bu.edu> The virtio-scsi fuzz target sets up and fuzzes the available virtio-scsi queues. After an element is placed on a queue, the fuzzer can select whether to perform a kick, or continue adding elements. Signed-off-by: Alexander Bulekov <alxndr@bu.edu> Reviewed-by: Darren Kenny <darren.kenny@oracle.com> Message-id: 20200220041118.23264-22-alxndr@bu.edu Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> --- tests/qtest/fuzz/Makefile.include | 1 + tests/qtest/fuzz/virtio_scsi_fuzz.c | 213 ++++++++++++++++++++++++++++ 2 files changed, 214 insertions(+) create mode 100644 tests/qtest/fuzz/virtio_scsi_fuzz.c diff --git a/tests/qtest/fuzz/Makefile.include b/tests/qtest/fuzz/Makefile.include index XXXXXXX..XXXXXXX 100644 --- a/tests/qtest/fuzz/Makefile.include +++ b/tests/qtest/fuzz/Makefile.include @@ -XXX,XX +XXX,XX @@ fuzz-obj-y += tests/qtest/fuzz/qos_fuzz.o # Targets fuzz-obj-y += tests/qtest/fuzz/i440fx_fuzz.o fuzz-obj-y += tests/qtest/fuzz/virtio_net_fuzz.o +fuzz-obj-y += tests/qtest/fuzz/virtio_scsi_fuzz.o FUZZ_CFLAGS += -I$(SRC_PATH)/tests -I$(SRC_PATH)/tests/qtest diff --git a/tests/qtest/fuzz/virtio_scsi_fuzz.c b/tests/qtest/fuzz/virtio_scsi_fuzz.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/tests/qtest/fuzz/virtio_scsi_fuzz.c @@ -XXX,XX +XXX,XX @@ +/* + * virtio-serial Fuzzing Target + * + * Copyright Red Hat Inc., 2019 + * + * Authors: + * Alexander Bulekov <alxndr@bu.edu> + * + * This work is licensed under the terms of the GNU GPL, version 2 or later. + * See the COPYING file in the top-level directory. + */ + +#include "qemu/osdep.h" + +#include "tests/qtest/libqtest.h" +#include "libqos/virtio-scsi.h" +#include "libqos/virtio.h" +#include "libqos/virtio-pci.h" +#include "standard-headers/linux/virtio_ids.h" +#include "standard-headers/linux/virtio_pci.h" +#include "standard-headers/linux/virtio_scsi.h" +#include "fuzz.h" +#include "fork_fuzz.h" +#include "qos_fuzz.h" + +#define PCI_SLOT 0x02 +#define PCI_FN 0x00 +#define QVIRTIO_SCSI_TIMEOUT_US (1 * 1000 * 1000) + +#define MAX_NUM_QUEUES 64 + +/* Based on tests/virtio-scsi-test.c */ +typedef struct { + int num_queues; + QVirtQueue *vq[MAX_NUM_QUEUES + 2]; +} QVirtioSCSIQueues; + +static QVirtioSCSIQueues *qvirtio_scsi_init(QVirtioDevice *dev, uint64_t mask) +{ + QVirtioSCSIQueues *vs; + uint64_t feat; + int i; + + vs = g_new0(QVirtioSCSIQueues, 1); + + feat = qvirtio_get_features(dev); + if (mask) { + feat &= ~QVIRTIO_F_BAD_FEATURE | mask; + } else { + feat &= ~(QVIRTIO_F_BAD_FEATURE | (1ull << VIRTIO_RING_F_EVENT_IDX)); + } + qvirtio_set_features(dev, feat); + + vs->num_queues = qvirtio_config_readl(dev, 0); + + for (i = 0; i < vs->num_queues + 2; i++) { + vs->vq[i] = qvirtqueue_setup(dev, fuzz_qos_alloc, i); + } + + qvirtio_set_driver_ok(dev); + + return vs; +} + +static void virtio_scsi_fuzz(QTestState *s, QVirtioSCSIQueues* queues, + const unsigned char *Data, size_t Size) +{ + /* + * Data is a sequence of random bytes. We split them up into "actions", + * followed by data: + * [vqa][dddddddd][vqa][dddd][vqa][dddddddddddd] ... + * The length of the data is specified by the preceding vqa.length + */ + typedef struct vq_action { + uint8_t queue; + uint8_t length; + uint8_t write; + uint8_t next; + uint8_t kick; + } vq_action; + + /* Keep track of the free head for each queue we interact with */ + bool vq_touched[MAX_NUM_QUEUES + 2] = {0}; + uint32_t free_head[MAX_NUM_QUEUES + 2]; + + QGuestAllocator *t_alloc = fuzz_qos_alloc; + + QVirtioSCSI *scsi = fuzz_qos_obj; + QVirtioDevice *dev = scsi->vdev; + QVirtQueue *q; + vq_action vqa; + while (Size >= sizeof(vqa)) { + /* Copy the action, so we can normalize length, queue and flags */ + memcpy(&vqa, Data, sizeof(vqa)); + + Data += sizeof(vqa); + Size -= sizeof(vqa); + + vqa.queue = vqa.queue % queues->num_queues; + /* Cap length at the number of remaining bytes in data */ + vqa.length = vqa.length >= Size ? Size : vqa.length; + vqa.write = vqa.write & 1; + vqa.next = vqa.next & 1; + vqa.kick = vqa.kick & 1; + + + q = queues->vq[vqa.queue]; + + /* Copy the data into ram, and place it on the virtqueue */ + uint64_t req_addr = guest_alloc(t_alloc, vqa.length); + qtest_memwrite(s, req_addr, Data, vqa.length); + if (vq_touched[vqa.queue] == 0) { + vq_touched[vqa.queue] = 1; + free_head[vqa.queue] = qvirtqueue_add(s, q, req_addr, vqa.length, + vqa.write, vqa.next); + } else { + qvirtqueue_add(s, q, req_addr, vqa.length, vqa.write , vqa.next); + } + + if (vqa.kick) { + qvirtqueue_kick(s, dev, q, free_head[vqa.queue]); + free_head[vqa.queue] = 0; + } + Data += vqa.length; + Size -= vqa.length; + } + /* In the end, kick each queue we interacted with */ + for (int i = 0; i < MAX_NUM_QUEUES + 2; i++) { + if (vq_touched[i]) { + qvirtqueue_kick(s, dev, queues->vq[i], free_head[i]); + } + } +} + +static void virtio_scsi_fork_fuzz(QTestState *s, + const unsigned char *Data, size_t Size) +{ + QVirtioSCSI *scsi = fuzz_qos_obj; + static QVirtioSCSIQueues *queues; + if (!queues) { + queues = qvirtio_scsi_init(scsi->vdev, 0); + } + if (fork() == 0) { + virtio_scsi_fuzz(s, queues, Data, Size); + flush_events(s); + _Exit(0); + } else { + wait(NULL); + } +} + +static void virtio_scsi_with_flag_fuzz(QTestState *s, + const unsigned char *Data, size_t Size) +{ + QVirtioSCSI *scsi = fuzz_qos_obj; + static QVirtioSCSIQueues *queues; + + if (fork() == 0) { + if (Size >= sizeof(uint64_t)) { + queues = qvirtio_scsi_init(scsi->vdev, *(uint64_t *)Data); + virtio_scsi_fuzz(s, queues, + Data + sizeof(uint64_t), Size - sizeof(uint64_t)); + flush_events(s); + } + _Exit(0); + } else { + wait(NULL); + } +} + +static void virtio_scsi_pre_fuzz(QTestState *s) +{ + qos_init_path(s); + counter_shm_init(); +} + +static void *virtio_scsi_test_setup(GString *cmd_line, void *arg) +{ + g_string_append(cmd_line, + " -drive file=blkdebug::null-co://," + "file.image.read-zeroes=on," + "if=none,id=dr1,format=raw,file.align=4k " + "-device scsi-hd,drive=dr1,lun=0,scsi-id=1"); + return arg; +} + + +static void register_virtio_scsi_fuzz_targets(void) +{ + fuzz_add_qos_target(&(FuzzTarget){ + .name = "virtio-scsi-fuzz", + .description = "Fuzz the virtio-scsi virtual queues, forking" + "for each fuzz run", + .pre_vm_init = &counter_shm_init, + .pre_fuzz = &virtio_scsi_pre_fuzz, + .fuzz = virtio_scsi_fork_fuzz,}, + "virtio-scsi", + &(QOSGraphTestOptions){.before = virtio_scsi_test_setup} + ); + + fuzz_add_qos_target(&(FuzzTarget){ + .name = "virtio-scsi-flags-fuzz", + .description = "Fuzz the virtio-scsi virtual queues, forking" + "for each fuzz run (also fuzzes the virtio flags)", + .pre_vm_init = &counter_shm_init, + .pre_fuzz = &virtio_scsi_pre_fuzz, + .fuzz = virtio_scsi_with_flag_fuzz,}, + "virtio-scsi", + &(QOSGraphTestOptions){.before = virtio_scsi_test_setup} + ); +} + +fuzz_target_init(register_virtio_scsi_fuzz_targets); -- 2.24.1 From: Alexander Bulekov <alxndr@bu.edu> Signed-off-by: Alexander Bulekov <alxndr@bu.edu> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Darren Kenny <darren.kenny@oracle.com> Message-id: 20200220041118.23264-23-alxndr@bu.edu Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> --- docs/devel/fuzzing.txt | 116 +++++++++++++++++++++++++++++++++++++++++ 1 file changed, 116 insertions(+) create mode 100644 docs/devel/fuzzing.txt diff --git a/docs/devel/fuzzing.txt b/docs/devel/fuzzing.txt new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/docs/devel/fuzzing.txt @@ -XXX,XX +XXX,XX @@ += Fuzzing = + +== Introduction == + +This document describes the virtual-device fuzzing infrastructure in QEMU and +how to use it to implement additional fuzzers. + +== Basics == + +Fuzzing operates by passing inputs to an entry point/target function. The +fuzzer tracks the code coverage triggered by the input. Based on these +findings, the fuzzer mutates the input and repeats the fuzzing. + +To fuzz QEMU, we rely on libfuzzer. Unlike other fuzzers such as AFL, libfuzzer +is an _in-process_ fuzzer. For the developer, this means that it is their +responsibility to ensure that state is reset between fuzzing-runs. + +== Building the fuzzers == + +NOTE: If possible, build a 32-bit binary. When forking, the 32-bit fuzzer is +much faster, since the page-map has a smaller size. This is due to the fact that +AddressSanitizer mmaps ~20TB of memory, as part of its detection. This results +in a large page-map, and a much slower fork(). + +To build the fuzzers, install a recent version of clang: +Configure with (substitute the clang binaries with the version you installed): + + CC=clang-8 CXX=clang++-8 /path/to/configure --enable-fuzzing + +Fuzz targets are built similarly to system/softmmu: + + make i386-softmmu/fuzz + +This builds ./i386-softmmu/qemu-fuzz-i386 + +The first option to this command is: --fuzz_taget=FUZZ_NAME +To list all of the available fuzzers run qemu-fuzz-i386 with no arguments. + +eg: + ./i386-softmmu/qemu-fuzz-i386 --fuzz-target=virtio-net-fork-fuzz + +Internally, libfuzzer parses all arguments that do not begin with "--". +Information about these is available by passing -help=1 + +Now the only thing left to do is wait for the fuzzer to trigger potential +crashes. + +== Adding a new fuzzer == +Coverage over virtual devices can be improved by adding additional fuzzers. +Fuzzers are kept in tests/qtest/fuzz/ and should be added to +tests/qtest/fuzz/Makefile.include + +Fuzzers can rely on both qtest and libqos to communicate with virtual devices. + +1. Create a new source file. For example ``tests/qtest/fuzz/foo-device-fuzz.c``. + +2. Write the fuzzing code using the libqtest/libqos API. See existing fuzzers +for reference. + +3. Register the fuzzer in ``tests/fuzz/Makefile.include`` by appending the +corresponding object to fuzz-obj-y + +Fuzzers can be more-or-less thought of as special qtest programs which can +modify the qtest commands and/or qtest command arguments based on inputs +provided by libfuzzer. Libfuzzer passes a byte array and length. Commonly the +fuzzer loops over the byte-array interpreting it as a list of qtest commands, +addresses, or values. + += Implementation Details = + +== The Fuzzer's Lifecycle == + +The fuzzer has two entrypoints that libfuzzer calls. libfuzzer provides it's +own main(), which performs some setup, and calls the entrypoints: + +LLVMFuzzerInitialize: called prior to fuzzing. Used to initialize all of the +necessary state + +LLVMFuzzerTestOneInput: called for each fuzzing run. Processes the input and +resets the state at the end of each run. + +In more detail: + +LLVMFuzzerInitialize parses the arguments to the fuzzer (must start with two +dashes, so they are ignored by libfuzzer main()). Currently, the arguments +select the fuzz target. Then, the qtest client is initialized. If the target +requires qos, qgraph is set up and the QOM/LIBQOS modules are initialized. +Then the QGraph is walked and the QEMU cmd_line is determined and saved. + +After this, the vl.c:qemu__main is called to set up the guest. There are +target-specific hooks that can be called before and after qemu_main, for +additional setup(e.g. PCI setup, or VM snapshotting). + +LLVMFuzzerTestOneInput: Uses qtest/qos functions to act based on the fuzz +input. It is also responsible for manually calling the main loop/main_loop_wait +to ensure that bottom halves are executed and any cleanup required before the +next input. + +Since the same process is reused for many fuzzing runs, QEMU state needs to +be reset at the end of each run. There are currently two implemented +options for resetting state: +1. Reboot the guest between runs. + Pros: Straightforward and fast for simple fuzz targets. + Cons: Depending on the device, does not reset all device state. If the + device requires some initialization prior to being ready for fuzzing + (common for QOS-based targets), this initialization needs to be done after + each reboot. + Example target: i440fx-qtest-reboot-fuzz +2. Run each test case in a separate forked process and copy the coverage + information back to the parent. This is fairly similar to AFL's "deferred" + fork-server mode [3] + Pros: Relatively fast. Devices only need to be initialized once. No need + to do slow reboots or vmloads. + Cons: Not officially supported by libfuzzer. Does not work well for devices + that rely on dedicated threads. + Example target: virtio-net-fork-fuzz -- 2.24.1