| Message ID | 1417598601-18537-2-git-send-email-mreitz@redhat.com |
|---|---|
| State | New |
| Headers | show |
On Wed, 12/03 10:23, Max Reitz wrote: > If vmdk blindly tries to use path_combine() using bs->file->filename as > the base file name, this will result in a bad error message for JSON > file names when calling bdrv_open(). It is better to only try > bs->file->exact_filename; if that is empty, bs->file->filename will be > useless for path_combine() and an error should be emitted (containing > bs->file->filename because bs->file->exact_filename is empty). > > Note that s->create_type does not need to be freed on error because it > will be freed by the caller (which ultimately is vmdk_open()). > > Signed-off-by: Max Reitz <mreitz@redhat.com> > --- > block/vmdk.c | 9 ++++++++- > 1 file changed, 8 insertions(+), 1 deletion(-) > > diff --git a/block/vmdk.c b/block/vmdk.c > index 2cbfd3e..fe549c2 100644 > --- a/block/vmdk.c > +++ b/block/vmdk.c > @@ -894,7 +894,14 @@ 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->filename, errp); > + > + if (!bs->file->exact_filename[0]) { > + error_setg(errp, "Cannot use extents with VMDK descriptor file '%s'", > + bs->file->filename); > + ret = -EINVAL; > + goto exit; > + } > + ret = vmdk_parse_extents(buf, bs, bs->file->exact_filename, errp); > exit: > return ret; > } > -- > 1.9.3 > Reviewed-by: Fam Zheng <famz@redhat.com>
Am 03.12.2014 um 10:23 hat Max Reitz geschrieben: > If vmdk blindly tries to use path_combine() using bs->file->filename as > the base file name, this will result in a bad error message for JSON > file names when calling bdrv_open(). It is better to only try > bs->file->exact_filename; if that is empty, bs->file->filename will be > useless for path_combine() and an error should be emitted (containing > bs->file->filename because bs->file->exact_filename is empty). > > Note that s->create_type does not need to be freed on error because it > will be freed by the caller (which ultimately is vmdk_open()). > > Signed-off-by: Max Reitz <mreitz@redhat.com> > --- > block/vmdk.c | 9 ++++++++- > 1 file changed, 8 insertions(+), 1 deletion(-) > > diff --git a/block/vmdk.c b/block/vmdk.c > index 2cbfd3e..fe549c2 100644 > --- a/block/vmdk.c > +++ b/block/vmdk.c > @@ -894,7 +894,14 @@ 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->filename, errp); > + > + if (!bs->file->exact_filename[0]) { > + error_setg(errp, "Cannot use extents with VMDK descriptor file '%s'", > + bs->file->filename); > + ret = -EINVAL; > + goto exit; > + } Isn't this overly restrictive? If the extent paths are all absolute, there is no reason not to open them. Or does the VMDK spec say that they are always relative? > + ret = vmdk_parse_extents(buf, bs, bs->file->exact_filename, errp); > exit: > return ret; > } Kevin
On 2014-12-03 at 13:44, Kevin Wolf wrote: > Am 03.12.2014 um 10:23 hat Max Reitz geschrieben: >> If vmdk blindly tries to use path_combine() using bs->file->filename as >> the base file name, this will result in a bad error message for JSON >> file names when calling bdrv_open(). It is better to only try >> bs->file->exact_filename; if that is empty, bs->file->filename will be >> useless for path_combine() and an error should be emitted (containing >> bs->file->filename because bs->file->exact_filename is empty). >> >> Note that s->create_type does not need to be freed on error because it >> will be freed by the caller (which ultimately is vmdk_open()). >> >> Signed-off-by: Max Reitz <mreitz@redhat.com> >> --- >> block/vmdk.c | 9 ++++++++- >> 1 file changed, 8 insertions(+), 1 deletion(-) >> >> diff --git a/block/vmdk.c b/block/vmdk.c >> index 2cbfd3e..fe549c2 100644 >> --- a/block/vmdk.c >> +++ b/block/vmdk.c >> @@ -894,7 +894,14 @@ 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->filename, errp); >> + >> + if (!bs->file->exact_filename[0]) { >> + error_setg(errp, "Cannot use extents with VMDK descriptor file '%s'", >> + bs->file->filename); >> + ret = -EINVAL; >> + goto exit; >> + } > Isn't this overly restrictive? If the extent paths are all absolute, > there is no reason not to open them. Or does the VMDK spec say that they > are always relative? Yes, you're right. Will respin. Max >> + ret = vmdk_parse_extents(buf, bs, bs->file->exact_filename, errp); >> exit: >> return ret; >> } > Kevin
diff --git a/block/vmdk.c b/block/vmdk.c index 2cbfd3e..fe549c2 100644 --- a/block/vmdk.c +++ b/block/vmdk.c @@ -894,7 +894,14 @@ 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->filename, errp); + + if (!bs->file->exact_filename[0]) { + error_setg(errp, "Cannot use extents with VMDK descriptor file '%s'", + bs->file->filename); + ret = -EINVAL; + goto exit; + } + ret = vmdk_parse_extents(buf, bs, bs->file->exact_filename, errp); exit: return ret; }
If vmdk blindly tries to use path_combine() using bs->file->filename as the base file name, this will result in a bad error message for JSON file names when calling bdrv_open(). It is better to only try bs->file->exact_filename; if that is empty, bs->file->filename will be useless for path_combine() and an error should be emitted (containing bs->file->filename because bs->file->exact_filename is empty). Note that s->create_type does not need to be freed on error because it will be freed by the caller (which ultimately is vmdk_open()). Signed-off-by: Max Reitz <mreitz@redhat.com> --- block/vmdk.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-)