debugfs: convert gid= argument from decimal, not octal
PATCH 1/1 debugfs: convert gid= argument from decimal, not octal
This patch technically breaks userspace, but I suspect that anyone who actually used this flag would have encountered this brokenness, declared it lunacy, and already sent a patch.
--- fs/debugfs/inode.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
fs/debugfs/inode.c +1 -1
@@ -176,7 +176,7 @@ static int debugfs_parse_options(char *data, struct debugfs_mount_opts *opts) opts->uid = uid;break ;case Opt_gid: -if (match_octal (&args[0 ], &option)) +if (match_int (&args[0 ], &option))return -EINVAL ; gid =make_kgid (current_user_ns (), option);if (!gid_valid (gid))
Signature · 1 line
1.8.0.3
Greg Kroah-Hartman
maintainerQuoted 3 lines · Dave Reisner
This patch technically breaks userspace, but I suspect that anyone who actually used this flag would have encountered this brokenness, declared it lunacy, and already sent a patch.
Kees is the one that originally did this change (I think, right?)
Kees, any objection to this patch? Will it break your existing systems?
thanks,
greg k-h
Quoted 21 lines
Signed-off-by Dave Reisner <[email protected]>--- fs/debugfs/inode.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)fs/debugfs/inode.c +1
-1@@ -176,7 +176,7 @@ static int debugfs_parse_options(char *data, struct debugfs_mount_opts *opts) opts->uid = uid;break ;case Opt_gid: -if (match_octal (&args[0 ], &option)) +if (match_int (&args[0 ], &option))return -EINVAL ; gid =make_kgid (current_user_ns (), option);if (!gid_valid (gid))Signature · 1 line
1.8.0.3
Quoted 6 lines · Greg Kroah-Hartman
Quoted 3 lines · Dave Reisner
This patch technically breaks userspace, but I suspect that anyone who actually used this flag would have encountered this brokenness, declared it lunacy, and already sent a patch.
Kees is the one that originally did this change (I think, right?)
I recommended the default change, but I think it was Vasiliy that added the gid option, IIRC.
Quoted 1 line
Kees, any objection to this patch? Will it break your existing systems?
Regardless, I have no objection -- this should be int, not octal.
-Kees
Quoted 11 lines
thanks,
greg k-h
Quoted 6 lines
Signed-off-by Dave Reisner <[email protected]>--- fs/debugfs/inode.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
Quoted 15 lines
Quoted 14 lines
fs/debugfs/inode.c +1
-2@@ -176,7 +176,7 @@ static int debugfs_parse_options(char *data, struct debugfs_mount_opts *opts) opts->uid = uid;break ;case Opt_gid: -if (match_octal (&args[0 ], &option)) +if (match_int (&args[0 ], &option))return -EINVAL ; gid =make_kgid (current_user_ns (), option);if (!gid_valid (gid)) --Quoted 1 line
1.8.0.3
--
Kees Cook
Chrome OS Security
Greg Kroah-Hartman
maintainerQuoted 11 lines · Kees Cook
Quoted 6 lines · Greg Kroah-Hartman
Quoted 3 lines · Dave Reisner
This patch technically breaks userspace, but I suspect that anyone who actually used this flag would have encountered this brokenness, declared it lunacy, and already sent a patch.
Kees is the one that originally did this change (I think, right?)
I recommended the default change, but I think it was Vasiliy that added the gid option, IIRC.
Ah, you are right, sorry about that.
Quoted 3 lines
Quoted 1 line
Kees, any objection to this patch? Will it break your existing systems?
Regardless, I have no objection -- this should be int, not octal.
Ok, I'll queue this up when I get back from vacation next week.
thanks,
greg k-h
Vasiliy Kulikov
Reviewed-byHi,
(CC'ed kernel-hardening ML)
Quoted 11 lines · Kees Cook
Quoted 6 lines · Greg Kroah-Hartman
Quoted 3 lines · Dave Reisner
This patch technically breaks userspace, but I suspect that anyone who actually used this flag would have encountered this brokenness, declared it lunacy, and already sent a patch.
Kees is the one that originally did this change (I think, right?)
I recommended the default change, but I think it was Vasiliy that added the gid option, IIRC.
Wow. It was not me :-) IIRC, I haven't proposed debugfs uid/gid stuff after Greg NACK'ed my sysfs uid/gid/umask mount options patch. IIRC, there were talks that debugfs must not be mounted on production systems because it exists for debug reasons only, so no additional security settings are needed.
My version of the patch (which was not posted on LKML):
http://www.openwall.com/lists/kernel-hardening/2011/06/05/4 (FWIW, it doesn't contain the bug in question.)
Quoted 4 lines
Quoted 2 lines
Kees, any objection to this patch? Will it break your existing systems?
Regardless, I have no objection -- this should be int, not octal.
No objection. GID should be decimal.
Quoted 9 lines
-Kees
Quoted 7 lines
thanks,
greg k-h
Quoted 2 lines
Signed-off-by Dave Reisner <[email protected]>
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
--- fs/debugfs/inode.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
Thanks,
Signature · 2 lines
Vasily Kulikov http://www.openwall.com - bringing security into open computing environments
Quoted 5 lines · Greg Kroah-Hartman
Quoted 3 lines · Dave Reisner
This patch technically breaks userspace, but I suspect that anyone who actually used this flag would have encountered this brokenness, declared it lunacy, and already sent a patch.
Quoted 8 lines
Kees is the one that originally did this change (I think, right?)
Kees, any objection to this patch? Will it break your existing systems?
thanks,
greg k-h
Actually the original patch was provided by Ludwig Nussel (cc'd) and commited as d6e486868cde58. Sorry, I should have mentioned at least the commit that introduced this in the original mailing.
d
Quoted 21 lines
Quoted 21 lines
Signed-off-by Dave Reisner <[email protected]>--- fs/debugfs/inode.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)fs/debugfs/inode.c +1
-1@@ -176,7 +176,7 @@ static int debugfs_parse_options(char *data, struct debugfs_mount_opts *opts) opts->uid = uid;break ;case Opt_gid: -if (match_octal (&args[0 ], &option)) +if (match_int (&args[0 ], &option))return -EINVAL ; gid =make_kgid (current_user_ns (), option);if (!gid_valid (gid))Signature · 1 line
1.8.0.3