* [PATCH 1/2] Btrfs-progs: set string end sing '\0' for property
@ 2014-04-25 6:29 Liu Bo
2014-04-25 6:29 ` [PATCH 2/2] Btrfs-progs: open file with O_RDONLY on getting property Liu Bo
` (2 more replies)
0 siblings, 3 replies; 8+ messages in thread
From: Liu Bo @ 2014-04-25 6:29 UTC (permalink / raw)
To: linux-btrfs; +Cc: Filipe David Borba Manana
Set string "xattr_name" 's end with '\0' so that it won't be
violated in memory.
With this fix, xfstest/btrfs/048 can pass on my box.
Signed-off-by: Liu Bo <bo.li.liu@oracle.com>
---
props.c | 1 +
1 file changed, 1 insertion(+)
diff --git a/props.c b/props.c
index 4d0aeea..53223a3 100644
--- a/props.c
+++ b/props.c
@@ -135,6 +135,7 @@ static int prop_compression(enum prop_object_type type,
}
memcpy(xattr_name, XATTR_BTRFS_PREFIX, XATTR_BTRFS_PREFIX_LEN);
memcpy(xattr_name + XATTR_BTRFS_PREFIX_LEN, name, strlen(name));
+ xattr_name[XATTR_BTRFS_PREFIX_LEN + strlen(name)] = '\0';
if (value)
sret = fsetxattr(fd, xattr_name, value, strlen(value), 0);
--
1.8.2.1
^ permalink raw reply related [flat|nested] 8+ messages in thread* [PATCH 2/2] Btrfs-progs: open file with O_RDONLY on getting property 2014-04-25 6:29 [PATCH 1/2] Btrfs-progs: set string end sing '\0' for property Liu Bo @ 2014-04-25 6:29 ` Liu Bo 2014-04-25 10:04 ` Filipe David Manana 2014-04-25 10:07 ` [PATCH 1/2] Btrfs-progs: set string end sing '\0' for property Filipe David Manana 2014-04-25 10:44 ` [PATCH v2] " Liu Bo 2 siblings, 1 reply; 8+ messages in thread From: Liu Bo @ 2014-04-25 6:29 UTC (permalink / raw) To: linux-btrfs; +Cc: Filipe David Borba Manana We don't need change an object when getting its property, so O_RDWR is not necessary in this case. Moreover, the object may be readonly, with O_RDWR it will fail. Signed-off-by: Liu Bo <bo.li.liu@oracle.com> --- props.c | 5 ++++- utils.c | 14 ++++++++++++-- utils.h | 1 + 3 files changed, 17 insertions(+), 3 deletions(-) diff --git a/props.c b/props.c index 53223a3..a49f8e0 100644 --- a/props.c +++ b/props.c @@ -120,7 +120,10 @@ static int prop_compression(enum prop_object_type type, char *buf = NULL; char *xattr_name = NULL; - fd = open_file_or_dir(object, &dirstream); + if (!value) + fd = open_file_or_dir_ro(object, &dirstream); + else + fd = open_file_or_dir(object, &dirstream); if (fd == -1) { ret = -errno; fprintf(stderr, "ERROR: open %s failed. %s\n", diff --git a/utils.c b/utils.c index 3e9c527..110a0d1 100644 --- a/utils.c +++ b/utils.c @@ -1622,7 +1622,7 @@ u64 parse_size(char *s) return strtoull(s, NULL, 10) * mult; } -int open_file_or_dir(const char *fname, DIR **dirstream) +static int __open_file_or_dir(const char *fname, DIR **dirstream, int ro) { int ret; struct stat st; @@ -1638,7 +1638,7 @@ int open_file_or_dir(const char *fname, DIR **dirstream) return -1; fd = dirfd(*dirstream); } else if (S_ISREG(st.st_mode) || S_ISLNK(st.st_mode)) { - fd = open(fname, O_RDWR); + fd = open(fname, (!!ro) ? O_RDONLY : O_RDWR); } else { /* * we set this on purpose, in case the caller output @@ -1655,6 +1655,16 @@ int open_file_or_dir(const char *fname, DIR **dirstream) return fd; } +int open_file_or_dir(const char *fname, DIR **dirstream) +{ + return __open_file_or_dir(fname, dirstream, 0); +} + +int open_file_or_dir_ro(const char *fname, DIR **dirstream) +{ + return __open_file_or_dir(fname, dirstream, 1); +} + void close_file_or_dir(int fd, DIR *dirstream) { if (dirstream) diff --git a/utils.h b/utils.h index 3c62066..5489cc2 100644 --- a/utils.h +++ b/utils.h @@ -72,6 +72,7 @@ int btrfs_scan_block_devices(int run_ioctl); u64 parse_size(char *s); u64 arg_strtou64(const char *str); int open_file_or_dir(const char *fname, DIR **dirstream); +int open_file_or_dir_ro(const char *fname, DIR **dirstream); void close_file_or_dir(int fd, DIR *dirstream); int get_fs_info(char *path, struct btrfs_ioctl_fs_info_args *fi_args, struct btrfs_ioctl_dev_info_args **di_ret); -- 1.8.2.1 ^ permalink raw reply related [flat|nested] 8+ messages in thread
* Re: [PATCH 2/2] Btrfs-progs: open file with O_RDONLY on getting property 2014-04-25 6:29 ` [PATCH 2/2] Btrfs-progs: open file with O_RDONLY on getting property Liu Bo @ 2014-04-25 10:04 ` Filipe David Manana 2014-04-25 10:38 ` Liu Bo 0 siblings, 1 reply; 8+ messages in thread From: Filipe David Manana @ 2014-04-25 10:04 UTC (permalink / raw) To: Liu Bo; +Cc: linux-btrfs@vger.kernel.org On Fri, Apr 25, 2014 at 7:29 AM, Liu Bo <bo.li.liu@oracle.com> wrote: > We don't need change an object when getting its property, > so O_RDWR is not necessary in this case. Moreover, the object > may be readonly, with O_RDWR it will fail. This was already addressed here Liu: https://patchwork.kernel.org/patch/3995151/ thanks > > Signed-off-by: Liu Bo <bo.li.liu@oracle.com> > --- > props.c | 5 ++++- > utils.c | 14 ++++++++++++-- > utils.h | 1 + > 3 files changed, 17 insertions(+), 3 deletions(-) > > diff --git a/props.c b/props.c > index 53223a3..a49f8e0 100644 > --- a/props.c > +++ b/props.c > @@ -120,7 +120,10 @@ static int prop_compression(enum prop_object_type type, > char *buf = NULL; > char *xattr_name = NULL; > > - fd = open_file_or_dir(object, &dirstream); > + if (!value) > + fd = open_file_or_dir_ro(object, &dirstream); > + else > + fd = open_file_or_dir(object, &dirstream); > if (fd == -1) { > ret = -errno; > fprintf(stderr, "ERROR: open %s failed. %s\n", > diff --git a/utils.c b/utils.c > index 3e9c527..110a0d1 100644 > --- a/utils.c > +++ b/utils.c > @@ -1622,7 +1622,7 @@ u64 parse_size(char *s) > return strtoull(s, NULL, 10) * mult; > } > > -int open_file_or_dir(const char *fname, DIR **dirstream) > +static int __open_file_or_dir(const char *fname, DIR **dirstream, int ro) > { > int ret; > struct stat st; > @@ -1638,7 +1638,7 @@ int open_file_or_dir(const char *fname, DIR **dirstream) > return -1; > fd = dirfd(*dirstream); > } else if (S_ISREG(st.st_mode) || S_ISLNK(st.st_mode)) { > - fd = open(fname, O_RDWR); > + fd = open(fname, (!!ro) ? O_RDONLY : O_RDWR); > } else { > /* > * we set this on purpose, in case the caller output > @@ -1655,6 +1655,16 @@ int open_file_or_dir(const char *fname, DIR **dirstream) > return fd; > } > > +int open_file_or_dir(const char *fname, DIR **dirstream) > +{ > + return __open_file_or_dir(fname, dirstream, 0); > +} > + > +int open_file_or_dir_ro(const char *fname, DIR **dirstream) > +{ > + return __open_file_or_dir(fname, dirstream, 1); > +} > + > void close_file_or_dir(int fd, DIR *dirstream) > { > if (dirstream) > diff --git a/utils.h b/utils.h > index 3c62066..5489cc2 100644 > --- a/utils.h > +++ b/utils.h > @@ -72,6 +72,7 @@ int btrfs_scan_block_devices(int run_ioctl); > u64 parse_size(char *s); > u64 arg_strtou64(const char *str); > int open_file_or_dir(const char *fname, DIR **dirstream); > +int open_file_or_dir_ro(const char *fname, DIR **dirstream); > void close_file_or_dir(int fd, DIR *dirstream); > int get_fs_info(char *path, struct btrfs_ioctl_fs_info_args *fi_args, > struct btrfs_ioctl_dev_info_args **di_ret); > -- > 1.8.2.1 > -- Filipe David Manana, "Reasonable men adapt themselves to the world. Unreasonable men adapt the world to themselves. That's why all progress depends on unreasonable men." ^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [PATCH 2/2] Btrfs-progs: open file with O_RDONLY on getting property 2014-04-25 10:04 ` Filipe David Manana @ 2014-04-25 10:38 ` Liu Bo 0 siblings, 0 replies; 8+ messages in thread From: Liu Bo @ 2014-04-25 10:38 UTC (permalink / raw) To: Filipe David Manana; +Cc: linux-btrfs@vger.kernel.org On Fri, Apr 25, 2014 at 11:04:39AM +0100, Filipe David Manana wrote: > On Fri, Apr 25, 2014 at 7:29 AM, Liu Bo <bo.li.liu@oracle.com> wrote: > > We don't need change an object when getting its property, > > so O_RDWR is not necessary in this case. Moreover, the object > > may be readonly, with O_RDWR it will fail. > > > This was already addressed here Liu: > https://patchwork.kernel.org/patch/3995151/ Okay, please ignore this. thanks, -liubo > > thanks > > > > > Signed-off-by: Liu Bo <bo.li.liu@oracle.com> > > --- > > props.c | 5 ++++- > > utils.c | 14 ++++++++++++-- > > utils.h | 1 + > > 3 files changed, 17 insertions(+), 3 deletions(-) > > > > diff --git a/props.c b/props.c > > index 53223a3..a49f8e0 100644 > > --- a/props.c > > +++ b/props.c > > @@ -120,7 +120,10 @@ static int prop_compression(enum prop_object_type type, > > char *buf = NULL; > > char *xattr_name = NULL; > > > > - fd = open_file_or_dir(object, &dirstream); > > + if (!value) > > + fd = open_file_or_dir_ro(object, &dirstream); > > + else > > + fd = open_file_or_dir(object, &dirstream); > > if (fd == -1) { > > ret = -errno; > > fprintf(stderr, "ERROR: open %s failed. %s\n", > > diff --git a/utils.c b/utils.c > > index 3e9c527..110a0d1 100644 > > --- a/utils.c > > +++ b/utils.c > > @@ -1622,7 +1622,7 @@ u64 parse_size(char *s) > > return strtoull(s, NULL, 10) * mult; > > } > > > > -int open_file_or_dir(const char *fname, DIR **dirstream) > > +static int __open_file_or_dir(const char *fname, DIR **dirstream, int ro) > > { > > int ret; > > struct stat st; > > @@ -1638,7 +1638,7 @@ int open_file_or_dir(const char *fname, DIR **dirstream) > > return -1; > > fd = dirfd(*dirstream); > > } else if (S_ISREG(st.st_mode) || S_ISLNK(st.st_mode)) { > > - fd = open(fname, O_RDWR); > > + fd = open(fname, (!!ro) ? O_RDONLY : O_RDWR); > > } else { > > /* > > * we set this on purpose, in case the caller output > > @@ -1655,6 +1655,16 @@ int open_file_or_dir(const char *fname, DIR **dirstream) > > return fd; > > } > > > > +int open_file_or_dir(const char *fname, DIR **dirstream) > > +{ > > + return __open_file_or_dir(fname, dirstream, 0); > > +} > > + > > +int open_file_or_dir_ro(const char *fname, DIR **dirstream) > > +{ > > + return __open_file_or_dir(fname, dirstream, 1); > > +} > > + > > void close_file_or_dir(int fd, DIR *dirstream) > > { > > if (dirstream) > > diff --git a/utils.h b/utils.h > > index 3c62066..5489cc2 100644 > > --- a/utils.h > > +++ b/utils.h > > @@ -72,6 +72,7 @@ int btrfs_scan_block_devices(int run_ioctl); > > u64 parse_size(char *s); > > u64 arg_strtou64(const char *str); > > int open_file_or_dir(const char *fname, DIR **dirstream); > > +int open_file_or_dir_ro(const char *fname, DIR **dirstream); > > void close_file_or_dir(int fd, DIR *dirstream); > > int get_fs_info(char *path, struct btrfs_ioctl_fs_info_args *fi_args, > > struct btrfs_ioctl_dev_info_args **di_ret); > > -- > > 1.8.2.1 > > > > > > -- > Filipe David Manana, > > "Reasonable men adapt themselves to the world. > Unreasonable men adapt the world to themselves. > That's why all progress depends on unreasonable men." ^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [PATCH 1/2] Btrfs-progs: set string end sing '\0' for property 2014-04-25 6:29 [PATCH 1/2] Btrfs-progs: set string end sing '\0' for property Liu Bo 2014-04-25 6:29 ` [PATCH 2/2] Btrfs-progs: open file with O_RDONLY on getting property Liu Bo @ 2014-04-25 10:07 ` Filipe David Manana 2014-04-25 10:37 ` Liu Bo 2014-04-25 10:44 ` [PATCH v2] " Liu Bo 2 siblings, 1 reply; 8+ messages in thread From: Filipe David Manana @ 2014-04-25 10:07 UTC (permalink / raw) To: Liu Bo; +Cc: linux-btrfs@vger.kernel.org On Fri, Apr 25, 2014 at 7:29 AM, Liu Bo <bo.li.liu@oracle.com> wrote: > Set string "xattr_name" 's end with '\0' so that it won't be > violated in memory. > > With this fix, xfstest/btrfs/048 can pass on my box. > > Signed-off-by: Liu Bo <bo.li.liu@oracle.com> > --- > props.c | 1 + > 1 file changed, 1 insertion(+) > > diff --git a/props.c b/props.c > index 4d0aeea..53223a3 100644 > --- a/props.c > +++ b/props.c > @@ -135,6 +135,7 @@ static int prop_compression(enum prop_object_type type, > } > memcpy(xattr_name, XATTR_BTRFS_PREFIX, XATTR_BTRFS_PREFIX_LEN); > memcpy(xattr_name + XATTR_BTRFS_PREFIX_LEN, name, strlen(name)); > + xattr_name[XATTR_BTRFS_PREFIX_LEN + strlen(name)] = '\0'; Buffer overrun. You need to sum + 1 to the malloc argument above. Thanks Liu. > > if (value) > sret = fsetxattr(fd, xattr_name, value, strlen(value), 0); > -- > 1.8.2.1 > -- Filipe David Manana, "Reasonable men adapt themselves to the world. Unreasonable men adapt the world to themselves. That's why all progress depends on unreasonable men." ^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [PATCH 1/2] Btrfs-progs: set string end sing '\0' for property 2014-04-25 10:07 ` [PATCH 1/2] Btrfs-progs: set string end sing '\0' for property Filipe David Manana @ 2014-04-25 10:37 ` Liu Bo 0 siblings, 0 replies; 8+ messages in thread From: Liu Bo @ 2014-04-25 10:37 UTC (permalink / raw) To: Filipe David Manana; +Cc: linux-btrfs@vger.kernel.org On Fri, Apr 25, 2014 at 11:07:49AM +0100, Filipe David Manana wrote: > On Fri, Apr 25, 2014 at 7:29 AM, Liu Bo <bo.li.liu@oracle.com> wrote: > > Set string "xattr_name" 's end with '\0' so that it won't be > > violated in memory. > > > > With this fix, xfstest/btrfs/048 can pass on my box. > > > > Signed-off-by: Liu Bo <bo.li.liu@oracle.com> > > --- > > props.c | 1 + > > 1 file changed, 1 insertion(+) > > > > diff --git a/props.c b/props.c > > index 4d0aeea..53223a3 100644 > > --- a/props.c > > +++ b/props.c > > @@ -135,6 +135,7 @@ static int prop_compression(enum prop_object_type type, > > } > > memcpy(xattr_name, XATTR_BTRFS_PREFIX, XATTR_BTRFS_PREFIX_LEN); > > memcpy(xattr_name + XATTR_BTRFS_PREFIX_LEN, name, strlen(name)); > > + xattr_name[XATTR_BTRFS_PREFIX_LEN + strlen(name)] = '\0'; > > Buffer overrun. You need to sum + 1 to the malloc argument above. Oops, I made that mistake. thanks, -liubo ^ permalink raw reply [flat|nested] 8+ messages in thread
* [PATCH v2] Btrfs-progs: set string end sing '\0' for property 2014-04-25 6:29 [PATCH 1/2] Btrfs-progs: set string end sing '\0' for property Liu Bo 2014-04-25 6:29 ` [PATCH 2/2] Btrfs-progs: open file with O_RDONLY on getting property Liu Bo 2014-04-25 10:07 ` [PATCH 1/2] Btrfs-progs: set string end sing '\0' for property Filipe David Manana @ 2014-04-25 10:44 ` Liu Bo 2014-04-25 10:51 ` Filipe David Manana 2 siblings, 1 reply; 8+ messages in thread From: Liu Bo @ 2014-04-25 10:44 UTC (permalink / raw) To: linux-btrfs; +Cc: Filipe David Manana Set string "xattr_name" 's end with '\0' so that it won't be violated in memory. With this fix, xfstest/btrfs/048 can pass on my box. Signed-off-by: Liu Bo <bo.li.liu@oracle.com> --- v2: avoid buffer overflow of malloc(). props.c | 3 ++- 1 files changed, 2 insertions(+), 1 deletions(-) diff --git a/props.c b/props.c index 9ff63ca..cc87454 100644 --- a/props.c +++ b/props.c @@ -131,13 +131,14 @@ static int prop_compression(enum prop_object_type type, goto out; } - xattr_name = malloc(XATTR_BTRFS_PREFIX_LEN + strlen(name)); + xattr_name = malloc(XATTR_BTRFS_PREFIX_LEN + strlen(name) + 1); if (!xattr_name) { ret = -ENOMEM; goto out; } memcpy(xattr_name, XATTR_BTRFS_PREFIX, XATTR_BTRFS_PREFIX_LEN); memcpy(xattr_name + XATTR_BTRFS_PREFIX_LEN, name, strlen(name)); + xattr_name[XATTR_BTRFS_PREFIX_LEN + strlen(name)] = '\0'; if (value) sret = fsetxattr(fd, xattr_name, value, strlen(value), 0); -- 1.7.7 ^ permalink raw reply related [flat|nested] 8+ messages in thread
* Re: [PATCH v2] Btrfs-progs: set string end sing '\0' for property 2014-04-25 10:44 ` [PATCH v2] " Liu Bo @ 2014-04-25 10:51 ` Filipe David Manana 0 siblings, 0 replies; 8+ messages in thread From: Filipe David Manana @ 2014-04-25 10:51 UTC (permalink / raw) To: Liu Bo; +Cc: linux-btrfs@vger.kernel.org On Fri, Apr 25, 2014 at 11:44 AM, Liu Bo <bo.li.liu@oracle.com> wrote: > Set string "xattr_name" 's end with '\0' so that it won't be > violated in memory. > > With this fix, xfstest/btrfs/048 can pass on my box. > > Signed-off-by: Liu Bo <bo.li.liu@oracle.com> Reviewed-by: Filipe Manana <fdmanana@gmail.com> > --- > v2: avoid buffer overflow of malloc(). > > props.c | 3 ++- > 1 files changed, 2 insertions(+), 1 deletions(-) > > diff --git a/props.c b/props.c > index 9ff63ca..cc87454 100644 > --- a/props.c > +++ b/props.c > @@ -131,13 +131,14 @@ static int prop_compression(enum prop_object_type type, > goto out; > } > > - xattr_name = malloc(XATTR_BTRFS_PREFIX_LEN + strlen(name)); > + xattr_name = malloc(XATTR_BTRFS_PREFIX_LEN + strlen(name) + 1); > if (!xattr_name) { > ret = -ENOMEM; > goto out; > } > memcpy(xattr_name, XATTR_BTRFS_PREFIX, XATTR_BTRFS_PREFIX_LEN); > memcpy(xattr_name + XATTR_BTRFS_PREFIX_LEN, name, strlen(name)); > + xattr_name[XATTR_BTRFS_PREFIX_LEN + strlen(name)] = '\0'; > > if (value) > sret = fsetxattr(fd, xattr_name, value, strlen(value), 0); > -- > 1.7.7 > -- Filipe David Manana, "Reasonable men adapt themselves to the world. Unreasonable men adapt the world to themselves. That's why all progress depends on unreasonable men." ^ permalink raw reply [flat|nested] 8+ messages in thread
end of thread, other threads:[~2014-04-25 10:51 UTC | newest] Thread overview: 8+ messages (download: mbox.gz follow: Atom feed -- links below jump to the message on this page -- 2014-04-25 6:29 [PATCH 1/2] Btrfs-progs: set string end sing '\0' for property Liu Bo 2014-04-25 6:29 ` [PATCH 2/2] Btrfs-progs: open file with O_RDONLY on getting property Liu Bo 2014-04-25 10:04 ` Filipe David Manana 2014-04-25 10:38 ` Liu Bo 2014-04-25 10:07 ` [PATCH 1/2] Btrfs-progs: set string end sing '\0' for property Filipe David Manana 2014-04-25 10:37 ` Liu Bo 2014-04-25 10:44 ` [PATCH v2] " Liu Bo 2014-04-25 10:51 ` Filipe David Manana
This is a public inbox, see mirroring instructions for how to clone and mirror all data and code used for this inbox