[PATCH 0/7] fixes for fsck.cramfs and other things

[PATCH 0/7] fixes for fsck.cramfs and other things

[Ruediger Meier]

From: Ruediger Meier <ruediger.meier@ga-group.nl>

Ruediger Meier (6):
  Revert "fsck.cramfs: Fix bus error on broken file system."
  fsck.cramfs: fix crash when superblock size is too small
  tests: rewrite fsck-bad-header
  fsck.cramfs: fix error message
  tests: fix fincore/count KNOWN_FAIL
  rfkill: provide RFKILL_TYPE_FM if undefined

root (1):
  tests: suppress warning for old sync(1)

 disk-utils/fsck.cramfs.c                          | 20 +++---
 sys-utils/rfkill.c                                |  5 ++
 tests/expected/cramfs/fsck-bad-header             |  2 -
 tests/expected/cramfs/fsck-bad-header-nopad-4K-be | 29 +++++++++
 tests/expected/cramfs/fsck-bad-header-nopad-4K-le | 29 +++++++++
 tests/expected/cramfs/fsck-bad-header-pad-4K-be   | 29 +++++++++
 tests/expected/cramfs/fsck-bad-header-pad-4K-le   | 29 +++++++++
 tests/expected/cramfs/fsck-bad-header-pad-64K-be  | 29 +++++++++
 tests/expected/cramfs/fsck-bad-header-pad-64K-le  | 29 +++++++++
 tests/functions.sh                                |  4 +-
 tests/ts/cramfs/fsck-bad-header                   | 76 +++++++++++++++++++++--
 tests/ts/fincore/count                            |  7 +--
 12 files changed, 266 insertions(+), 22 deletions(-)
 delete mode 100644 tests/expected/cramfs/fsck-bad-header
 create mode 100644 tests/expected/cramfs/fsck-bad-header-nopad-4K-be
 create mode 100644 tests/expected/cramfs/fsck-bad-header-nopad-4K-le
 create mode 100644 tests/expected/cramfs/fsck-bad-header-pad-4K-be
 create mode 100644 tests/expected/cramfs/fsck-bad-header-pad-4K-le
 create mode 100644 tests/expected/cramfs/fsck-bad-header-pad-64K-be
 create mode 100644 tests/expected/cramfs/fsck-bad-header-pad-64K-le

-- 
2.13.6

[PATCH 1/7] Revert "fsck.cramfs: Fix bus error on broken file system."

[Ruediger Meier]

From: Ruediger Meier <ruediger.meier@ga-group.nl>

This reverts commit 7cb962c77015e9383b53eeb22ce732cb5216bbc3.

It can't be right that we mmap (start + super.size) bytes from a file
which is usually only super.size bytes large. The patch "fixed" a
problem when super.size is bad but now it fails for the correct case:

$ mkdir -p root/subdir
$ ./mkfs.cramfs -p root cramfs
$ ./fsck.cramfs cramfs
Bus error (core dumped)

We will fix the original problem later.

CC: Tobias Stoeckmann <tobias@stoeckmann.org>
Signed-off-by: Ruediger Meier <ruediger.meier@ga-group.nl>
---
 disk-utils/fsck.cramfs.c | 13 ++++++-------
 1 file changed, 6 insertions(+), 7 deletions(-)

diff --git a/disk-utils/fsck.cramfs.c b/disk-utils/fsck.cramfs.c
index b2a3cc153..50c7d33b9 100644
--- a/disk-utils/fsck.cramfs.c
+++ b/disk-utils/fsck.cramfs.c
@@ -220,24 +220,23 @@ static void test_crc(int start)
 	crc = crc32(0L, NULL, 0);
 
 	buf =
-	    mmap(NULL, start + super.size, PROT_READ | PROT_WRITE, MAP_PRIVATE, fd, 0);
+	    mmap(NULL, super.size, PROT_READ | PROT_WRITE, MAP_PRIVATE, fd, 0);
 	if (buf == MAP_FAILED) {
 		buf =
-		    mmap(NULL, start + super.size, PROT_READ | PROT_WRITE,
+		    mmap(NULL, super.size, PROT_READ | PROT_WRITE,
 			 MAP_PRIVATE | MAP_ANONYMOUS, -1, 0);
 		if (buf != MAP_FAILED) {
-			if (lseek(fd, start, SEEK_SET) == (off_t) -1)
+			if (lseek(fd, 0, SEEK_SET) == (off_t) -1)
 				err(FSCK_EX_ERROR, _("seek on %s failed"), filename);
-			if (read(fd, (unsigned char *) buf + start, super.size) !=
-			    (ssize_t) super.size)
+			if (read(fd, buf, super.size) != (ssize_t) super.size)
 				err(FSCK_EX_ERROR, _("cannot read %s"), filename);
 		}
 	}
 	if (buf != MAP_FAILED) {
 		((struct cramfs_super *)((unsigned char *) buf + start))->fsid.crc =
 		    crc32(0L, NULL, 0);
-		crc = crc32(crc, (unsigned char *) buf + start, super.size);
-		munmap(buf, start + super.size);
+		crc = crc32(crc, (unsigned char *) buf + start, super.size - start);
+		munmap(buf, super.size);
 	} else {
 		int retval;
 		size_t length = 0;
-- 
2.13.6

[PATCH 2/7] fsck.cramfs: fix crash when superblock size is too small

[Ruediger Meier]

From: Ruediger Meier <ruediger.meier@ga-group.nl>

This hopefully fixes the original problem addressed by the reverted
patch 7cb962c7.

The bug was introduced by myself in
    f991dbd3 "fsck.cramfs: allow smaller superblock sizes"

CC: Tobias Stoeckmann <tobias@stoeckmann.org>
Signed-off-by: Ruediger Meier <ruediger.meier@ga-group.nl>
---
 disk-utils/fsck.cramfs.c              | 2 +-
 tests/expected/cramfs/fsck-bad-header | 3 +--
 tests/ts/cramfs/fsck-bad-header       | 2 +-
 3 files changed, 3 insertions(+), 4 deletions(-)

diff --git a/disk-utils/fsck.cramfs.c b/disk-utils/fsck.cramfs.c
index 50c7d33b9..820816b14 100644
--- a/disk-utils/fsck.cramfs.c
+++ b/disk-utils/fsck.cramfs.c
@@ -192,7 +192,7 @@ static void test_super(int *start, size_t * length)
 		errx(FSCK_EX_ERROR, _("unsupported filesystem features"));
 
 	/* What are valid superblock sizes? */
-	if (super.size < sizeof(struct cramfs_super))
+	if (super.size < *start + sizeof(struct cramfs_super))
 		errx(FSCK_EX_UNCORRECTED, _("superblock size (%d) too small"),
 		     super.size);
 
diff --git a/tests/expected/cramfs/fsck-bad-header b/tests/expected/cramfs/fsck-bad-header
index ade8a9c3a..41c2c125e 100644
--- a/tests/expected/cramfs/fsck-bad-header
+++ b/tests/expected/cramfs/fsck-bad-header
@@ -1,2 +1 @@
-fsck.cramfs: file extends past end of filesystem
-fsck.cramfs: crc error
+fsck.cramfs: superblock size (76) too small
diff --git a/tests/ts/cramfs/fsck-bad-header b/tests/ts/cramfs/fsck-bad-header
index 8d5309c35..add388577 100755
--- a/tests/ts/cramfs/fsck-bad-header
+++ b/tests/ts/cramfs/fsck-bad-header
@@ -16,7 +16,7 @@
 # GNU General Public License for more details.
 #
 TS_TOPDIR="${0%/*}/../.."
-TS_DESC="fsck endianness"
+TS_DESC="fsck bad header"
 
 . $TS_TOPDIR/functions.sh
 ts_init "$*"
-- 
2.13.6

[PATCH 3/7] tests: rewrite fsck-bad-header

[Ruediger Meier]

From: Ruediger Meier <ruediger.meier@ga-group.nl>

Fix test for systems with pagesize != 4096
Loop over many combinations of sizes, endianness and blocksizes.

Signed-off-by: Ruediger Meier <ruediger.meier@ga-group.nl>
---
 tests/expected/cramfs/fsck-bad-header             |  1 -
 tests/expected/cramfs/fsck-bad-header-nopad-4K-be | 29 +++++++++
 tests/expected/cramfs/fsck-bad-header-nopad-4K-le | 29 +++++++++
 tests/expected/cramfs/fsck-bad-header-pad-4K-be   | 29 +++++++++
 tests/expected/cramfs/fsck-bad-header-pad-4K-le   | 29 +++++++++
 tests/expected/cramfs/fsck-bad-header-pad-64K-be  | 29 +++++++++
 tests/expected/cramfs/fsck-bad-header-pad-64K-le  | 29 +++++++++
 tests/ts/cramfs/fsck-bad-header                   | 74 +++++++++++++++++++++--
 8 files changed, 243 insertions(+), 6 deletions(-)
 delete mode 100644 tests/expected/cramfs/fsck-bad-header
 create mode 100644 tests/expected/cramfs/fsck-bad-header-nopad-4K-be
 create mode 100644 tests/expected/cramfs/fsck-bad-header-nopad-4K-le
 create mode 100644 tests/expected/cramfs/fsck-bad-header-pad-4K-be
 create mode 100644 tests/expected/cramfs/fsck-bad-header-pad-4K-le
 create mode 100644 tests/expected/cramfs/fsck-bad-header-pad-64K-be
 create mode 100644 tests/expected/cramfs/fsck-bad-header-pad-64K-le

diff --git a/tests/expected/cramfs/fsck-bad-header b/tests/expected/cramfs/fsck-bad-header
deleted file mode 100644
index 41c2c125e..000000000
--- a/tests/expected/cramfs/fsck-bad-header
+++ /dev/null
@@ -1 +0,0 @@
-fsck.cramfs: superblock size (76) too small
diff --git a/tests/expected/cramfs/fsck-bad-header-nopad-4K-be b/tests/expected/cramfs/fsck-bad-header-nopad-4K-be
new file mode 100644
index 000000000..397d1ba29
--- /dev/null
+++ b/tests/expected/cramfs/fsck-bad-header-nopad-4K-be
@@ -0,0 +1,29 @@
+## size: 0
+fsck.cramfs: superblock size (0) too small
+ret: 4
+
+## size: 75
+fsck.cramfs: superblock size (75) too small
+ret: 4
+
+## size: 76
+fsck.cramfs: file extends past end of filesystem
+fsck.cramfs: crc error
+ret: 4
+
+## size: 4095
+fsck.cramfs: file extends past end of filesystem
+fsck.cramfs: crc error
+ret: 4
+
+## size: 4096
+ret: 0
+
+## size: 4097
+fsck.cramfs: file length too short
+ret: 4
+
+## size: 4294967295
+fsck.cramfs: file length too short
+ret: 4
+
diff --git a/tests/expected/cramfs/fsck-bad-header-nopad-4K-le b/tests/expected/cramfs/fsck-bad-header-nopad-4K-le
new file mode 100644
index 000000000..397d1ba29
--- /dev/null
+++ b/tests/expected/cramfs/fsck-bad-header-nopad-4K-le
@@ -0,0 +1,29 @@
+## size: 0
+fsck.cramfs: superblock size (0) too small
+ret: 4
+
+## size: 75
+fsck.cramfs: superblock size (75) too small
+ret: 4
+
+## size: 76
+fsck.cramfs: file extends past end of filesystem
+fsck.cramfs: crc error
+ret: 4
+
+## size: 4095
+fsck.cramfs: file extends past end of filesystem
+fsck.cramfs: crc error
+ret: 4
+
+## size: 4096
+ret: 0
+
+## size: 4097
+fsck.cramfs: file length too short
+ret: 4
+
+## size: 4294967295
+fsck.cramfs: file length too short
+ret: 4
+
diff --git a/tests/expected/cramfs/fsck-bad-header-pad-4K-be b/tests/expected/cramfs/fsck-bad-header-pad-4K-be
new file mode 100644
index 000000000..8e2316183
--- /dev/null
+++ b/tests/expected/cramfs/fsck-bad-header-pad-4K-be
@@ -0,0 +1,29 @@
+## size: 76
+fsck.cramfs: superblock size (76) too small
+ret: 4
+
+## size: 587
+fsck.cramfs: superblock size (587) too small
+ret: 4
+
+## size: 588
+fsck.cramfs: file extends past end of filesystem
+fsck.cramfs: crc error
+ret: 4
+
+## size: 4095
+fsck.cramfs: file extends past end of filesystem
+fsck.cramfs: crc error
+ret: 4
+
+## size: 4096
+ret: 0
+
+## size: 4097
+fsck.cramfs: file length too short
+ret: 4
+
+## size: 4294967295
+fsck.cramfs: file length too short
+ret: 4
+
diff --git a/tests/expected/cramfs/fsck-bad-header-pad-4K-le b/tests/expected/cramfs/fsck-bad-header-pad-4K-le
new file mode 100644
index 000000000..8e2316183
--- /dev/null
+++ b/tests/expected/cramfs/fsck-bad-header-pad-4K-le
@@ -0,0 +1,29 @@
+## size: 76
+fsck.cramfs: superblock size (76) too small
+ret: 4
+
+## size: 587
+fsck.cramfs: superblock size (587) too small
+ret: 4
+
+## size: 588
+fsck.cramfs: file extends past end of filesystem
+fsck.cramfs: crc error
+ret: 4
+
+## size: 4095
+fsck.cramfs: file extends past end of filesystem
+fsck.cramfs: crc error
+ret: 4
+
+## size: 4096
+ret: 0
+
+## size: 4097
+fsck.cramfs: file length too short
+ret: 4
+
+## size: 4294967295
+fsck.cramfs: file length too short
+ret: 4
+
diff --git a/tests/expected/cramfs/fsck-bad-header-pad-64K-be b/tests/expected/cramfs/fsck-bad-header-pad-64K-be
new file mode 100644
index 000000000..15d4ff190
--- /dev/null
+++ b/tests/expected/cramfs/fsck-bad-header-pad-64K-be
@@ -0,0 +1,29 @@
+## size: 76
+fsck.cramfs: superblock size (76) too small
+ret: 4
+
+## size: 587
+fsck.cramfs: superblock size (587) too small
+ret: 4
+
+## size: 588
+fsck.cramfs: file extends past end of filesystem
+fsck.cramfs: crc error
+ret: 4
+
+## size: 65535
+fsck.cramfs: file extends past end of filesystem
+fsck.cramfs: crc error
+ret: 4
+
+## size: 65536
+ret: 0
+
+## size: 65537
+fsck.cramfs: file length too short
+ret: 4
+
+## size: 4294967295
+fsck.cramfs: file length too short
+ret: 4
+
diff --git a/tests/expected/cramfs/fsck-bad-header-pad-64K-le b/tests/expected/cramfs/fsck-bad-header-pad-64K-le
new file mode 100644
index 000000000..15d4ff190
--- /dev/null
+++ b/tests/expected/cramfs/fsck-bad-header-pad-64K-le
@@ -0,0 +1,29 @@
+## size: 76
+fsck.cramfs: superblock size (76) too small
+ret: 4
+
+## size: 587
+fsck.cramfs: superblock size (587) too small
+ret: 4
+
+## size: 588
+fsck.cramfs: file extends past end of filesystem
+fsck.cramfs: crc error
+ret: 4
+
+## size: 65535
+fsck.cramfs: file extends past end of filesystem
+fsck.cramfs: crc error
+ret: 4
+
+## size: 65536
+ret: 0
+
+## size: 65537
+fsck.cramfs: file length too short
+ret: 4
+
+## size: 4294967295
+fsck.cramfs: file length too short
+ret: 4
+
diff --git a/tests/ts/cramfs/fsck-bad-header b/tests/ts/cramfs/fsck-bad-header
index add388577..c71782bde 100755
--- a/tests/ts/cramfs/fsck-bad-header
+++ b/tests/ts/cramfs/fsck-bad-header
@@ -25,14 +25,78 @@ ts_check_test_command "$TS_CMD_MKCRAMFS"
 ts_check_test_command "$TS_CMD_FSCKCRAMFS"
 ts_check_prog "dd"
 
-IMAGE_SOURCE="$TS_OUTDIR/${TS_TESTNAME}-data/root"
+function num2binary()
+{
+	local num=$1
+	local endian=$2
+
+	test "$num" -ge 0 -a "$num" -le 4294967295 || return 1
+	test "$endian" = "be" -o "$endian" = "le" || return 1
+
+	# how to do that easier?
+	if test "$endian" = "be"; then
+		echo -en "$(printf "%08x" "$1" | sed 's/\(..\)/\\x\1/g')"
+	else
+		echo -en "$(printf "%08x" "$1" | sed 's/^\(..\)\(..\)\(..\)\(..\)$/\\x\4\\x\3\\x\2\\x\1/')"
+	fi
+}
+
+function fsck_loop_sizes()
+{
+	local endian=$1  # be, le
+	local seek=$2    # 4 for nopad, 516 for pad
+	shift 2          # the rest are sizes to loop over
+
+	for size in "$@"; do
+		ts_log "## size: $size"
+		cp -a "$IMAGE_FILE" "$IMAGE_FILE.tmp"
+		num2binary "$size" $endian |
+			dd of="$IMAGE_FILE.tmp" bs=1 seek="$seek" count=4 conv=notrunc &> /dev/null
+		$TS_CMD_FSCKCRAMFS "$IMAGE_FILE.tmp" >> $TS_OUTPUT 2>&1
+		ts_log "ret: $?
+"
+	done
+	rm -f "$IMAGE_FILE"
+}
+
+
+IMAGE_SOURCE="$TS_OUTDIR/${TS_TESTNAME}-data"
 IMAGE_FILE="$TS_OUTDIR/${TS_TESTNAME}-cramfs.img"
 
 mkdir -p "${IMAGE_SOURCE}/subdir" &> /dev/null
-$TS_CMD_MKCRAMFS -p -N big $IMAGE_SOURCE $IMAGE_FILE &> /dev/null
-echo -ne \\00\\x4c |
-	dd of=$IMAGE_FILE bs=1 seek=518 count=2 conv=notrunc &> /dev/null
-$TS_CMD_FSCKCRAMFS $IMAGE_FILE >> $TS_OUTPUT 2>&1
+
+ts_init_subtest "nopad-4K-be"
+$TS_CMD_MKCRAMFS -N big -b 4096 $IMAGE_SOURCE $IMAGE_FILE &> /dev/null
+fsck_loop_sizes be 4  0 75 76 4095 4096 4097 4294967295
+rm -f "$IMAGE_FILE"
+ts_finalize_subtest
+
+ts_init_subtest "nopad-4K-le"
+$TS_CMD_MKCRAMFS -N little -b 4096 $IMAGE_SOURCE $IMAGE_FILE &> /dev/null
+fsck_loop_sizes le 4  0 75 76 4095 4096 4097 4294967295
+ts_finalize_subtest
+
+ts_init_subtest "pad-4K-be"
+$TS_CMD_MKCRAMFS -p -N big -b 4096 $IMAGE_SOURCE $IMAGE_FILE &> /dev/null
+fsck_loop_sizes be 516  76 587 588 4095 4096 4097 4294967295
+ts_finalize_subtest
+
+ts_init_subtest "pad-4K-le"
+$TS_CMD_MKCRAMFS -p -N little -b 4096 $IMAGE_SOURCE $IMAGE_FILE &> /dev/null
+fsck_loop_sizes le 516  76 587 588 4095 4096 4097 4294967295
+ts_finalize_subtest
+
+ts_init_subtest "pad-64K-be"
+$TS_CMD_MKCRAMFS -p -N big -b 65536 $IMAGE_SOURCE $IMAGE_FILE &> /dev/null
+fsck_loop_sizes be 516  76 587 588 65535 65536 65537 4294967295
+ts_finalize_subtest
+
+ts_init_subtest "pad-64K-le"
+$TS_CMD_MKCRAMFS -p -N little -b 65536 $IMAGE_SOURCE $IMAGE_FILE &> /dev/null
+fsck_loop_sizes le 516  76 587 588 65535 65536 65537 4294967295
+ts_finalize_subtest
+
+rm -rf "$IMAGE_SOURCE" "$IMAGE_FILE.tmp"
 
 ts_finalize
 
-- 
2.13.6

[PATCH 4/7] fsck.cramfs: fix error message

[Ruediger Meier]

From: Ruediger Meier <ruediger.meier@ga-group.nl>

errno is invalid in case that we just read(3) too short.

Signed-off-by: Ruediger Meier <ruediger.meier@ga-group.nl>
---
 disk-utils/fsck.cramfs.c | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)

diff --git a/disk-utils/fsck.cramfs.c b/disk-utils/fsck.cramfs.c
index 820816b14..2a4f75a8c 100644
--- a/disk-utils/fsck.cramfs.c
+++ b/disk-utils/fsck.cramfs.c
@@ -226,10 +226,15 @@ static void test_crc(int start)
 		    mmap(NULL, super.size, PROT_READ | PROT_WRITE,
 			 MAP_PRIVATE | MAP_ANONYMOUS, -1, 0);
 		if (buf != MAP_FAILED) {
+			ssize_t tmp;
 			if (lseek(fd, 0, SEEK_SET) == (off_t) -1)
 				err(FSCK_EX_ERROR, _("seek on %s failed"), filename);
-			if (read(fd, buf, super.size) != (ssize_t) super.size)
+			tmp = read(fd, buf, super.size);
+			if (tmp < 0)
 				err(FSCK_EX_ERROR, _("cannot read %s"), filename);
+			if (tmp != (ssize_t) super.size)
+				errx(FSCK_EX_ERROR, _("failed to read %"PRIu32" bytes from file %s"),
+					super.size, filename);
 		}
 	}
 	if (buf != MAP_FAILED) {
-- 
2.13.6

[PATCH 5/7] tests: suppress warning for old sync(1)

[Ruediger Meier]

From: root <root@adsiz>

Old sync(1) prints a warning which looks ugly among our
test output: "sync: ignoring all arguments"

Seen on travis, Ubuntu <= 14.04 (Trusty).

Signed-off-by: Ruediger Meier <ruediger.meier@ga-group.nl>
---
 tests/functions.sh | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/tests/functions.sh b/tests/functions.sh
index ef11a893e..d1c97e0b6 100644
--- a/tests/functions.sh
+++ b/tests/functions.sh
@@ -635,7 +635,7 @@ function ts_fstab_open {
 
 function ts_fstab_close {
 	echo "# -->" >> /etc/fstab
-	sync /etc/fstab
+	sync /etc/fstab 2>/dev/null
 }
 
 function ts_fstab_addline {
@@ -669,7 +669,7 @@ function ts_fstab_clean {
 s/# <!-- util-linux.*-->//;
 /^$/d" /etc/fstab
 
-	sync /etc/fstab
+	sync /etc/fstab 2>/dev/null
 	ts_unlock "fstab"
 }
 
-- 
2.13.6

[PATCH 6/7] tests: fix fincore/count KNOWN_FAIL

[Ruediger Meier]

From: Ruediger Meier <ruediger.meier@ga-group.nl>

It has to be set after calling ts_init.

Signed-off-by: Ruediger Meier <ruediger.meier@ga-group.nl>
---
 tests/ts/fincore/count | 7 +++----
 1 file changed, 3 insertions(+), 4 deletions(-)

diff --git a/tests/ts/fincore/count b/tests/ts/fincore/count
index c2dff7e34..46700a32e 100755
--- a/tests/ts/fincore/count
+++ b/tests/ts/fincore/count
@@ -3,15 +3,14 @@
 TS_TOPDIR="${0%/*}/../.."
 TS_DESC="count file contents in core"
 
-# Send patch if you know how to keep it portable and robust. Thanks.
-TS_KNOWN_FAIL="yes"
-
-
 . $TS_TOPDIR/functions.sh
 ts_init "$*"
 
 ts_check_test_command "$TS_HELPER_SYSINFO"
 
+# Send patch if you know how to keep it portable and robust. Thanks.
+TS_KNOWN_FAIL="yes"
+
 function header
 {
     echo "[" "$1" "]"
-- 
2.13.6

[PATCH 7/7] rfkill: provide RFKILL_TYPE_FM if undefined

[Ruediger Meier]

From: Ruediger Meier <ruediger.meier@ga-group.nl>

As discussed last year it's nice to be compatible to 2.6.32
https://www.spinics.net/lists/util-linux-ng/msg13963.html

BTW also re-define NUM_RFKILL_TYPES if needed, although we are
not really using it.

Signed-off-by: Ruediger Meier <ruediger.meier@ga-group.nl>
---
 sys-utils/rfkill.c | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/sys-utils/rfkill.c b/sys-utils/rfkill.c
index 75804ad41..031fe436f 100644
--- a/sys-utils/rfkill.c
+++ b/sys-utils/rfkill.c
@@ -45,7 +45,12 @@
  * year 2009 (2.6.33) or older.
  */
 #ifndef RFKILL_TYPE_NFC
+# ifndef RFKILL_TYPE_FM
+#  define RFKILL_TYPE_FM	RFKILL_TYPE_GPS + 1
+# endif
 # define RFKILL_TYPE_NFC	RFKILL_TYPE_FM + 1
+# undef NUM_RFKILL_TYPES
+# define NUM_RFKILL_TYPES	RFKILL_TYPE_NFC + 1
 #endif
 
 struct rfkill_type_str {
-- 
2.13.6

Re: [PATCH 2/7] fsck.cramfs: fix crash when superblock size is too small

[Tobias Stöckmann]

> This hopefully fixes the original problem addressed by the reverted
> patch 7cb962c7.

It does.

>  	/* What are valid superblock sizes? */
> -	if (super.size < sizeof(struct cramfs_super))
> +	if (super.size < *start + sizeof(struct cramfs_super))

And this is a more elegant (and working) solution. I was not sure how
much adjustment this if-statement can handle due to the comment.


Thanks and sorry for the regression,

Tobias

Re: [PATCH 0/7] fixes for fsck.cramfs and other things

[Karel Zak]

On Tue, Jan 23, 2018 at 04:59:21PM +0100, Ruediger Meier wrote:
> From: Ruediger Meier <ruediger.meier@ga-group.nl>
> 
> Ruediger Meier (6):
>   Revert "fsck.cramfs: Fix bus error on broken file system."
>   fsck.cramfs: fix crash when superblock size is too small
>   tests: rewrite fsck-bad-header

Nice.

>   fsck.cramfs: fix error message
>   tests: fix fincore/count KNOWN_FAIL
>   rfkill: provide RFKILL_TYPE_FM if undefined

Applied, thanks!

    Karel


-- 
 Karel Zak  <kzak@redhat.com>
 http://karelzak.blogspot.com

Re: [PATCH 2/7] fsck.cramfs: fix crash when superblock size is too small

[Ruediger Meier]

On Tuesday 23 January 2018, Tobias Stc3b6ckmann wrote:
> > This hopefully fixes the original problem addressed by the reverted
> > patch 7cb962c7.
>
> It does.
>
> >  	/* What are valid superblock sizes? */
> > -	if (super.size < sizeof(struct cramfs_super))
> > +	if (super.size < *start + sizeof(struct cramfs_super))
>
> And this is a more elegant (and working) solution. I was not sure how
> much adjustment this if-statement can handle due to the comment.

The comment is a bit stupid. Actually super.size should be a multiple of 
PAGE_SIZE of the running kernel to be able to mount it on the current 
system. I've relaxed this requirement to be able to fsck a cramfs from 
other systems. This is IMO a nice feature because in real life cramfs 
will be created often on cross-compiler hosts to be deployed to another 
embedded system. Though maybe we should add a warning like "This image 
is not mountable on the current system".

>
> Thanks and sorry for the regression,

All fine, you discovered a bug I made. :)

> Tobias
> --
> To unsubscribe from this list: send the line "unsubscribe util-linux"
> in the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH 0/7] fixes for fsck.cramfs and other things

[Ruediger Meier]

On Wednesday 24 January 2018, Karel Zak wrote:
> On Tue, Jan 23, 2018 at 04:59:21PM +0100, Ruediger Meier wrote:
> > From: Ruediger Meier <ruediger.meier@ga-group.nl>
> >
> > Ruediger Meier (6):
> >   Revert "fsck.cramfs: Fix bus error on broken file system."
> >   fsck.cramfs: fix crash when superblock size is too small
> >   tests: rewrite fsck-bad-header
>
> Nice.
>
> >   fsck.cramfs: fix error message

BTW while we are at this. I wonder why we have 3 different ways 
(fallbacks) in fsck.cramfs.c test_crc() how to read the image.

1. mmap
2. read into anonymous mmap
3. read blockwise

Does anybody knows why is this and whether we still need all these 
fallbacks?

cu,
Rudi

Re: [PATCH 0/7] fixes for fsck.cramfs and other things

[Karel Zak]

On Wed, Jan 24, 2018 at 03:47:57PM +0100, Ruediger Meier wrote:
> On Wednesday 24 January 2018, Karel Zak wrote:
> > On Tue, Jan 23, 2018 at 04:59:21PM +0100, Ruediger Meier wrote:
> > > From: Ruediger Meier <ruediger.meier@ga-group.nl>
> > >
> > > Ruediger Meier (6):
> > >   Revert "fsck.cramfs: Fix bus error on broken file system."
> > >   fsck.cramfs: fix crash when superblock size is too small
> > >   tests: rewrite fsck-bad-header
> >
> > Nice.
> >
> > >   fsck.cramfs: fix error message
> 
> BTW while we are at this. I wonder why we have 3 different ways 
> (fallbacks) in fsck.cramfs.c test_crc() how to read the image.
> 
> 1. mmap
> 2. read into anonymous mmap
> 3. read blockwise
> 
> Does anybody knows why is this and whether we still need all these 
> fallbacks?

The question is why we need mmap at all here;-)

    Karel

-- 
 Karel Zak  <kzak@redhat.com>
 http://karelzak.blogspot.com