Re: O_TMPFILE fs corruption (Re: Linux 3.11-rc4)

From: Andrew Lutomirski
Date: Mon Aug 05 2013 - 00:45:43 EST


On Sun, Aug 4, 2013 at 8:45 PM, Linus Torvalds
<torvalds@xxxxxxxxxxxxxxxxxxxx> wrote:
> The patch looks right to me - we should pass in similar flags for the
> create case as for tmpfile to the filesystem.

Alternatively, in case anyone ever wants to add more O_TMPFILE-related
flags, open could return -EINVAL if __O_TMPFILE is set and mode
contains bits outside S_IALLUGO. Given the ABI discussion about
O_TMPFILE, this might be useful some day.

--Andy

>
> But let's make sure we're all on the same page. Al?
>
> Linus
>
> On Sun, Aug 4, 2013 at 7:34 PM, Andy Lutomirski <luto@xxxxxxx> wrote:
>> On 08/04/2013 02:09 PM, Linus Torvalds wrote:
>>>
>>> It's that time of the week again..
>>
>>
>> I still get filesystem corruption with O_TMPFILE. The program below, run as
>> flinktest foo proc (or flinktest foo linkat if you're root) will produce a
>> bogus inode. On ext4, once the inode is gone from cache, the inode will be
>> impossible to delete and will require a fsck to fix
>>
>> A patch (not necessarily the appropriate fix) is here:
>>
>> http://article.gmane.org/gmane.linux.kernel/1537088
>>
>> --- cut here ---
>>
>> #include <stdio.h>
>> #include <err.h>
>> #include <fcntl.h>
>> #include <unistd.h>
>> #include <string.h>
>>
>> #define __O_TMPFILE 020000000
>> #define O_TMPFILE (__O_TMPFILE | O_DIRECTORY)
>> #define AT_EMPTY_PATH 0x1000
>>
>> int main(int argc, char **argv)
>> {
>> char buf[128];
>>
>> if (argc != 3)
>> errx(1, "Usage: flinktest PATH linkat|proc");
>>
>> int fd = open(".", O_TMPFILE | O_RDWR, 0600);
>> if (fd == -1)
>> err(1, "O_TMPFILE");
>> write(fd, "test", 4);
>>
>> if (!strcmp(argv[2], "linkat")) {
>> if (linkat(fd, "", AT_FDCWD, argv[1], AT_EMPTY_PATH) != 0)
>> err(1, "linkat");
>> } else if (!strcmp(argv[2], "proc")) {
>> sprintf(buf, "/proc/self/fd/%d", fd);
>> if (linkat(AT_FDCWD, buf, AT_FDCWD, argv[1], AT_SYMLINK_FOLLOW) != 0)
>> err(1, "linkat");
>> } else {
>> errx(1, "invalid mode");
>> }
>> return 0;
>> }
>>
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/