| Message ID | 20210430221621.7497-1-pc@cjr.nz |
|---|---|
| State | Superseded |
| Headers | show |
| Series | cifs: fix regression when mounting shares with prefix paths | expand |
On Fri, 30 Apr 2021 19:16:21 -0300, Paulo Alcantara wrote: > The commit 315db9a05b7a ("cifs: fix leak in cifs_smb3_do_mount() ctx") > revealed an existing bug when mounting shares that contain a prefix > path or DFS links. Sorry for the mess. One question... ... > if (mntopts) { > char *ip; > > - cifs_dbg(FYI, "%s: mntopts=%s\n", __func__, mntopts); > rc = smb3_parse_opt(mntopts, "ip", &ip); > - if (!rc && !cifs_convert_address((struct sockaddr *)&ctx->dstaddr, ip, > - strlen(ip))) { > - cifs_dbg(VFS, "%s: failed to convert ip address\n", __func__); > - return -EINVAL; > + if (!rc) { > + rc = cifs_convert_address((struct sockaddr *)&ctx->dstaddr, ip, strlen(ip)); > + kfree(ip); > + if (!rc) { > + cifs_dbg(VFS, "%s: failed to convert ip address\n", __func__); > + return -EINVAL; > + } > } > } > > @@ -3189,7 +3198,7 @@ cifs_setup_volume_info(struct smb3_fs_context *ctx, const char *mntopts, const c > return -EINVAL; > } > > - return rc; > + return 0; > } It seems that smb3_parse_opt() errors will no longer be propagated here. Is that intentional? Cheers, David
David Disseldorp <ddiss@suse.de> writes: > On Fri, 30 Apr 2021 19:16:21 -0300, Paulo Alcantara wrote: > >> The commit 315db9a05b7a ("cifs: fix leak in cifs_smb3_do_mount() ctx") >> revealed an existing bug when mounting shares that contain a prefix >> path or DFS links. > > Sorry for the mess. One question... > > ... >> if (mntopts) { >> char *ip; >> >> - cifs_dbg(FYI, "%s: mntopts=%s\n", __func__, mntopts); >> rc = smb3_parse_opt(mntopts, "ip", &ip); >> - if (!rc && !cifs_convert_address((struct sockaddr *)&ctx->dstaddr, ip, >> - strlen(ip))) { >> - cifs_dbg(VFS, "%s: failed to convert ip address\n", __func__); >> - return -EINVAL; >> + if (!rc) { >> + rc = cifs_convert_address((struct sockaddr *)&ctx->dstaddr, ip, strlen(ip)); >> + kfree(ip); >> + if (!rc) { >> + cifs_dbg(VFS, "%s: failed to convert ip address\n", __func__); >> + return -EINVAL; >> + } >> } >> } >> >> @@ -3189,7 +3198,7 @@ cifs_setup_volume_info(struct smb3_fs_context *ctx, const char *mntopts, const c >> return -EINVAL; >> } >> >> - return rc; >> + return 0; >> } > > It seems that smb3_parse_opt() errors will no longer be propagated here. > Is that intentional? That was a mistake, actually. Will fix it in v2. Thanks!
On Mon, 3 May 2021 11:55:26 -0300, Paulo Alcantara wrote: > The commit 315db9a05b7a ("cifs: fix leak in cifs_smb3_do_mount() ctx") > revealed an existing bug when mounting shares that contain a prefix > path or DFS links. > > cifs_setup_volume_info() requires the @devname to contain the full > path (UNC + prefix) to update the fs context with the new UNC and > prepath values, however we were passing only the UNC > path (old_ctx->UNC) in @device thus discarding any prefix paths. > > Instead of concatenating both old_ctx->{UNC,prepath} and pass it in > @devname, just keep the dup'ed values of UNC and prepath in > cifs_sb->ctx after calling smb3_fs_context_dup(), and fix > smb3_parse_devname() to correctly parse and not leak the new UNC and > prefix paths. > > Cc: <stable@vger.kernel.org> # v5.11+ > Fixes: 315db9a05b7a ("cifs: fix leak in cifs_smb3_do_mount() ctx") > Signed-off-by: Paulo Alcantara (SUSE) <pc@cjr.nz> > --- > fs/cifs/cifsfs.c | 8 +------- > fs/cifs/connect.c | 24 ++++++++++++++++++------ > fs/cifs/fs_context.c | 4 ++++ > 3 files changed, 23 insertions(+), 13 deletions(-) > > diff --git a/fs/cifs/cifsfs.c b/fs/cifs/cifsfs.c > index 8a6894577697..d7ea9c5fe0f8 100644 > --- a/fs/cifs/cifsfs.c > +++ b/fs/cifs/cifsfs.c > @@ -863,13 +863,7 @@ cifs_smb3_do_mount(struct file_system_type *fs_type, > goto out; > } > > - /* cifs_setup_volume_info->smb3_parse_devname() redups UNC & prepath */ > - kfree(cifs_sb->ctx->UNC); > - cifs_sb->ctx->UNC = NULL; > - kfree(cifs_sb->ctx->prepath); > - cifs_sb->ctx->prepath = NULL; > - > - rc = cifs_setup_volume_info(cifs_sb->ctx, NULL, old_ctx->UNC); > + rc = cifs_setup_volume_info(cifs_sb->ctx, NULL, NULL); IIUC, with the new behaviour, this call becomes pretty much an if (!cifs_sb->ctx->username) root = ERR_PTR(-EINVAL); So it might be worth simplifying this further. Either way: Acked-by: David Disseldorp <ddiss@suse.de> Cheers, David
David Disseldorp <ddiss@suse.de> writes: > On Mon, 3 May 2021 11:55:26 -0300, Paulo Alcantara wrote: >> - rc = cifs_setup_volume_info(cifs_sb->ctx, NULL, old_ctx->UNC); >> + rc = cifs_setup_volume_info(cifs_sb->ctx, NULL, NULL); > > IIUC, with the new behaviour, this call becomes pretty much an > if (!cifs_sb->ctx->username) > root = ERR_PTR(-EINVAL); Yeah. We should get rid of the damn thing at some point. I've just tried to keep the logic we currently have in fs/cifs/connect.c by composing new options based upon the chased DFS referrals. Thanks for pointing it out!
added acked-by and merged into cifs-2.6.git for-next On Mon, May 3, 2021 at 6:20 PM David Disseldorp <ddiss@suse.de> wrote: > > On Mon, 3 May 2021 11:55:26 -0300, Paulo Alcantara wrote: > > > The commit 315db9a05b7a ("cifs: fix leak in cifs_smb3_do_mount() ctx") > > revealed an existing bug when mounting shares that contain a prefix > > path or DFS links. > > > > cifs_setup_volume_info() requires the @devname to contain the full > > path (UNC + prefix) to update the fs context with the new UNC and > > prepath values, however we were passing only the UNC > > path (old_ctx->UNC) in @device thus discarding any prefix paths. > > > > Instead of concatenating both old_ctx->{UNC,prepath} and pass it in > > @devname, just keep the dup'ed values of UNC and prepath in > > cifs_sb->ctx after calling smb3_fs_context_dup(), and fix > > smb3_parse_devname() to correctly parse and not leak the new UNC and > > prefix paths. > > > > Cc: <stable@vger.kernel.org> # v5.11+ > > Fixes: 315db9a05b7a ("cifs: fix leak in cifs_smb3_do_mount() ctx") > > Signed-off-by: Paulo Alcantara (SUSE) <pc@cjr.nz> > > --- > > fs/cifs/cifsfs.c | 8 +------- > > fs/cifs/connect.c | 24 ++++++++++++++++++------ > > fs/cifs/fs_context.c | 4 ++++ > > 3 files changed, 23 insertions(+), 13 deletions(-) > > > > diff --git a/fs/cifs/cifsfs.c b/fs/cifs/cifsfs.c > > index 8a6894577697..d7ea9c5fe0f8 100644 > > --- a/fs/cifs/cifsfs.c > > +++ b/fs/cifs/cifsfs.c > > @@ -863,13 +863,7 @@ cifs_smb3_do_mount(struct file_system_type *fs_type, > > goto out; > > } > > > > - /* cifs_setup_volume_info->smb3_parse_devname() redups UNC & prepath */ > > - kfree(cifs_sb->ctx->UNC); > > - cifs_sb->ctx->UNC = NULL; > > - kfree(cifs_sb->ctx->prepath); > > - cifs_sb->ctx->prepath = NULL; > > - > > - rc = cifs_setup_volume_info(cifs_sb->ctx, NULL, old_ctx->UNC); > > + rc = cifs_setup_volume_info(cifs_sb->ctx, NULL, NULL); > > IIUC, with the new behaviour, this call becomes pretty much an > if (!cifs_sb->ctx->username) > root = ERR_PTR(-EINVAL); > > So it might be worth simplifying this further. Either way: > Acked-by: David Disseldorp <ddiss@suse.de> > > Cheers, David -- Thanks, Steve
diff --git a/fs/cifs/cifsfs.c b/fs/cifs/cifsfs.c index 8a6894577697..d7ea9c5fe0f8 100644 --- a/fs/cifs/cifsfs.c +++ b/fs/cifs/cifsfs.c @@ -863,13 +863,7 @@ cifs_smb3_do_mount(struct file_system_type *fs_type, goto out; } - /* cifs_setup_volume_info->smb3_parse_devname() redups UNC & prepath */ - kfree(cifs_sb->ctx->UNC); - cifs_sb->ctx->UNC = NULL; - kfree(cifs_sb->ctx->prepath); - cifs_sb->ctx->prepath = NULL; - - rc = cifs_setup_volume_info(cifs_sb->ctx, NULL, old_ctx->UNC); + rc = cifs_setup_volume_info(cifs_sb->ctx, NULL, NULL); if (rc) { root = ERR_PTR(rc); goto out; diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c index becd5f807787..04a06e22e715 100644 --- a/fs/cifs/connect.c +++ b/fs/cifs/connect.c @@ -3159,19 +3159,28 @@ static int do_dfs_failover(const char *path, const char *full_path, struct cifs_ int cifs_setup_volume_info(struct smb3_fs_context *ctx, const char *mntopts, const char *devname) { - int rc = 0; + int rc; - smb3_parse_devname(devname, ctx); + if (devname) { + cifs_dbg(FYI, "%s: devname=%s\n", __func__, devname); + rc = smb3_parse_devname(devname, ctx); + if (rc) { + cifs_dbg(VFS, "%s: failed to parse %s: %d\n", __func__, devname, rc); + return rc; + } + } if (mntopts) { char *ip; - cifs_dbg(FYI, "%s: mntopts=%s\n", __func__, mntopts); rc = smb3_parse_opt(mntopts, "ip", &ip); - if (!rc && !cifs_convert_address((struct sockaddr *)&ctx->dstaddr, ip, - strlen(ip))) { - cifs_dbg(VFS, "%s: failed to convert ip address\n", __func__); - return -EINVAL; + if (!rc) { + rc = cifs_convert_address((struct sockaddr *)&ctx->dstaddr, ip, strlen(ip)); + kfree(ip); + if (!rc) { + cifs_dbg(VFS, "%s: failed to convert ip address\n", __func__); + return -EINVAL; + } } } @@ -3189,7 +3198,7 @@ cifs_setup_volume_info(struct smb3_fs_context *ctx, const char *mntopts, const c return -EINVAL; } - return rc; + return 0; } static int diff --git a/fs/cifs/fs_context.c b/fs/cifs/fs_context.c index 1d6e0e15b034..3bcf881c3ae9 100644 --- a/fs/cifs/fs_context.c +++ b/fs/cifs/fs_context.c @@ -476,6 +476,7 @@ smb3_parse_devname(const char *devname, struct smb3_fs_context *ctx) /* move "pos" up to delimiter or NULL */ pos += len; + kfree(ctx->UNC); ctx->UNC = kstrndup(devname, pos - devname, GFP_KERNEL); if (!ctx->UNC) return -ENOMEM; @@ -486,6 +487,9 @@ smb3_parse_devname(const char *devname, struct smb3_fs_context *ctx) if (*pos == '/' || *pos == '\\') pos++; + kfree(ctx->prepath); + ctx->prepath = NULL; + /* If pos is NULL then no prepath */ if (!*pos) return 0;
The commit 315db9a05b7a ("cifs: fix leak in cifs_smb3_do_mount() ctx") revealed an existing bug when mounting shares that contain a prefix path or DFS links. cifs_setup_volume_info() requires the @devname to contain the full path (UNC + prefix) to update the fs context with the new UNC and prepath values, however we were passing only the UNC path (old_ctx->UNC) in @device thus discarding any prefix paths. Instead of concatenating both old_ctx->{UNC,prepath} and pass it in @devname, just keep the dup'ed values of UNC and prepath in cifs_sb->ctx after calling smb3_fs_context_dup(), and fix smb3_parse_devname() to correctly parse and not leak the new UNC and prefix paths. Cc: <stable@vger.kernel.org> # v5.11+ Fixes: 315db9a05b7a ("cifs: fix leak in cifs_smb3_do_mount() ctx") Signed-off-by: Paulo Alcantara (SUSE) <pc@cjr.nz> --- fs/cifs/cifsfs.c | 8 +------- fs/cifs/connect.c | 25 +++++++++++++++++-------- fs/cifs/fs_context.c | 4 ++++ 3 files changed, 22 insertions(+), 15 deletions(-)