Skip to content

strbuf: simplify `strbuf_attach()` usage

Martin Ågren git 1 version In review

Get series with b4
b4 am [email protected]

Versions newest first

  1. v1 Latest6 patches 1 replyIn review

Patches in v1 No applies recorded yet

#Subject and filesSizeRepliesApplied
1/6am: use `strbuf_attach()` correctlybuiltin/am.c+1 -10 replies
2/6strbuf_attach: correctly pass in `strlen() + 1` for `alloc`mailinfo.c path.c refs/files-backend.c +1 more+5 -40 replies
3/6strbuf: use `strbuf_attach()` correctlystrbuf.c+1 -10 replies
4/6fast-import: avoid awkward use of `strbuf_attach()`fast-import.c+4 -30 replies
5/6rerere: avoid awkward use of `strbuf_attach()`rerere.c+2 -10 replies
6/6strbuf: simplify `strbuf_attach()` usageapply.c archive.c blame.c +11 more+29 -280 replies

Reviews trailers per patch

No Reviewed-by, Acked-by or Tested-by trailers for v1 yet.

Read the v1 thread

Cover letter [PATCH 0/6] strbuf: simplify `strbuf_attach()` usage
Quoted 8 lines · Martin Ågren

Review

Quoted 2 lines
(quoted diff) +1 -1
-       strbuf_attach(line, out, strlen(out), strlen(out));
+       strbuf_attach(line, out, out_len, out_len);

This conversion is ok as such. I wondered why we pass in the same value twice (before and after this patch). Turns out this usage is wrong (as per the documentation in strbuf.h) but safe (as per my understanding of the implementation in strbuf.c). I'll follow up with a series that fell out of that investigation.

Here's that series. It could go in parallel to Danh's, or one could go on top of the other. Any of those would be ok with me.

I think strbuf_attach() is sufficiently hard to use that we might as well simplify it for the majority of the users that don't care about the distinction between the string's length and the size of the allocated memory, and avoid it for the few remaining users that are just as well off using strbuf_add().

The summary is that this function takes len and alloc, where the latter must be greater than the former, yet several callers use the same value for both. I first thought this could cause quite hairy problems such as writing outside of allocated memory, but it doesn't look that way. See the patches for more information.

An alternative to the approach taken here would be to introduce strbuf_attachstr() and convert most existing users, then convert the few remaining ones to use the new function or to move in another direction. But the new name is a downer -- what else would you attach to a strbuf if not a string?

Another alternative would be to first convert certain users away from the function, then simplify it for the remainder. I preferred to first spend a few patches converting a few existing callers to make it clear that those are ok.

Martin

Martin Ågren (6):
am: use strbuf_attach() correctly
strbuf_attach: correctly pass in strlen() + 1 for alloc
strbuf: use strbuf_attach() correctly
fast-import: avoid awkward use of strbuf_attach()
rerere: avoid awkward use of strbuf_attach()
strbuf: simplify strbuf_attach() usage

 strbuf.h             | 16 ++++++++--------
 apply.c              |  2 +-
 archive.c            |  2 +-
 blame.c              |  2 +-
 builtin/am.c         |  2 +-
 convert.c            |  4 ++--
 fast-import.c        |  7 ++++---
 imap-send.c          |  2 +-
 mailinfo.c           |  2 +-
 merge-recursive.c    |  2 +-
 path.c               |  3 +--
 pretty.c             |  4 ++--
 refs/files-backend.c |  2 +-
 rerere.c             |  3 ++-
 strbuf.c             | 11 +++++++----
 trailer.c            |  2 +-
 16 files changed, 35 insertions(+), 31 deletions(-)
Signature · 1 line
2.26.1