git/list[1] front-page[2] threads[3] people[4] search[5] about
 

Re: [PATCH 8/7] read_pack_header: handle signed/unsigned comparison in read result

From
Jonathan Nieder <jrnieder@gmail.com>
Date
Sep 13, 2017, 19:11 UTC
Message-ID
<20170913191102.GF27425@aiede.mtv.corp.google.com>
In-Reply-To
<20170913184722.imcexldnufskugqz@sigill.intra.peff.net>
Jeff King wrote:
Show 24 quoted lines
> Subject: [PATCH] read_pack_header: handle signed/unsigned comparison in read  result
>
> The result of read_in_full() may be -1 if we saw an error.
> But in comparing it to a sizeof() result, that "-1" will be
> promoted to size_t. In fact, the largest possible size_t
> which is much bigger than our struct size. This means that
> our "< sizeof(header)" error check won't trigger.
>
> In practice, we'd go on to read uninitialized memory and
> compare it to the PACK signature, which is likely to fail.
> But we shouldn't get there.
>
> We can fix this by making a direct "!=" comparison to the
> requested size, rather than "<". This means that errors get
> lumped in with short reads, but that's sufficient for our
> purposes here. There's no PH_ERROR tp represent our case.
> And anyway, this function reads from pipes and network
> sockets. A network error may racily appear as EOF to us
> anyway if there's data left in the socket buffers.
>
> Signed-off-by: Jeff King <peff@peff.net>
> ---
>  sha1_file.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
This patch 8 (but not patches 2-7) is
Reviewed-by: Jonathan Nieder <jrnieder@gmail.com>
Thanks.
An alternative would be to use something like
	< (int) sizeof(*header)
to force it to be signed, but I think this is clearer.

Using the following semantic patch, I find this and the example from patch 1 and no others:

  @@
  expression fd;
  expression buf;
  expression len;
  size_t rhs;
  @@
  -read_in_full(fd, buf, len) < rhs
  +ERROR()
  @@
  expression fd;
  expression buf;
  expression len;
  size_t rhs;
  @@
  -write_in_full(fd, buf, len) < rhs
  +ERROR()

Thanks, Jonathan

Show 11 quoted lines
> --- a/sha1_file.c
> +++ b/sha1_file.c
> @@ -1850,7 +1850,7 @@ int index_path(struct object_id *oid, const char *path, struct stat *st, unsigne
>  
>  int read_pack_header(int fd, struct pack_header *header)
>  {
> -	if (read_in_full(fd, header, sizeof(*header)) < sizeof(*header))
> +	if (read_in_full(fd, header, sizeof(*header)) != sizeof(*header))
>  		/* "eof before pack header was fully read" */
>  		return PH_ERROR_EOF;
>  
Previous: Jeff King
Message 46 of 46 in “Bug: git branch --unset-upstream command can nuke config when disk is full.”
  1. demerphqSep 13, 2017
  2. Jeff KingSep 13, 2017
  3. demerphqSep 13, 2017
  4. Jeff KingSep 13, 2017
  5. demerphqSep 13, 2017
  6. Jeff KingSep 13, 2017
  7. demerphqSep 13, 2017
  8. Jeff KingSep 13, 2017
  9. demerphqSep 13, 2017
  10. 0/7 config.c may fail to notice some write() failuresJeff King, Sep 13, 2017
  11. 1/7 config: avoid "write_in_full(fd, buf, len) < len" patternJeff King, Sep 13, 2017
  12. Jonathan NiederSep 13, 2017
  13. Jeff KingSep 13, 2017
  14. Jonathan NiederSep 13, 2017
  15. Jeff KingSep 13, 2017
  16. config: avoid "write_in_full(fd, buf, len) < len" patternJeff King, Sep 13, 2017
  17. Jonathan NiederSep 13, 2017
  18. Jeff KingSep 13, 2017
  19. Jonathan NiederSep 13, 2017
  20. Jonathan NiederSep 13, 2017
  21. Ramsay JonesSep 13, 2017
  22. Ramsay JonesSep 13, 2017
  23. Jeff KingSep 15, 2017
  24. Ramsay JonesSep 15, 2017
  25. Junio C HamanoSep 13, 2017
  26. 2/7 get-tar-commit-id: check write_in_full() return against 0Jeff King, Sep 13, 2017
  27. Jonathan NiederSep 13, 2017
  28. Jeff KingSep 13, 2017
  29. Jeff KingSep 13, 2017
  30. Jonathan NiederSep 13, 2017
  31. Jeff KingSep 15, 2017
  32. 3/7 avoid "write_in_full(fd, buf, len) != len" patternJeff King, Sep 13, 2017
  33. Jonathan NiederSep 13, 2017
  34. Jeff KingSep 15, 2017
  35. 4/7 convert less-trivial versions of "write_in_full() != len"Jeff King, Sep 13, 2017
  36. Jonathan NiederSep 13, 2017
  37. 5/7 pkt-line: check write_in_full() errors against "< 0"Jeff King, Sep 13, 2017
  38. Jonathan NiederSep 13, 2017
  39. 6/7 notes-merge: use ssize_t for write_in_full() return valueJeff King, Sep 13, 2017
  40. Jonathan NiederSep 13, 2017
  41. Jeff KingSep 15, 2017
  42. 7/7 config: flip return value of store_write_*()Jeff King, Sep 13, 2017
  43. Jonathan NiederSep 13, 2017
  44. Jeff KingSep 15, 2017
  45. 8/7 read_pack_header: handle signed/unsigned comparison in read resultJeff King, Sep 13, 2017
  46. Jonathan NiederSep 13, 2017

Read the whole thread, see it on lore, or plain text.

$ cat FOOTERMessages come from the public archive at lore.kernel.org/git, fetched every hour. The front page is chosen and written each morning by an AI editor and can be wrong; the threads themselves are the record. About and API. For agents: an MCP server at https://gitlist.dev/mcp, and any thread, story or person page as Markdown by adding .md to its URL (or sending Accept: text/markdown). Details in /llms.txt.