Git development
 help / color / mirror / Atom feed
From: Jeff King <peff@peff.net>
To: Patrick Steinhardt <ps@pks.im>
Cc: git@vger.kernel.org, Phillip Wood <phillip.wood123@gmail.com>,
	Junio C Hamano <gitster@pobox.com>,
	Karthik Nayak <karthik.188@gmail.com>,
	Toon Claes <toon@iotcl.com>
Subject: Re: [PATCH v2 01/16] object-file: fix -Wsign-compare warnings
Date: Sun, 5 Apr 2026 02:56:02 -0400	[thread overview]
Message-ID: <20260405065602.GA1481314@coredump.intra.peff.net> (raw)
In-Reply-To: <20250717-pks-object-file-wo-the-repository-v2-1-36d2cd6c700e@pks.im>

On Thu, Jul 17, 2025 at 06:56:27AM +0200, Patrick Steinhardt wrote:

> @@ -1268,7 +1265,7 @@ int index_fd(struct index_state *istate, struct object_id *oid,
>  		ret = index_stream_convert_blob(istate, oid, fd, path, flags);
>  	else if (!S_ISREG(st->st_mode))
>  		ret = index_pipe(istate, oid, fd, type, path, flags);
> -	else if (st->st_size <= repo_settings_get_big_file_threshold(the_repository) ||
> +	else if ((st->st_size >= 0 && (size_t) st->st_size <= repo_settings_get_big_file_threshold(the_repository)) ||
>  		 type != OBJ_BLOB ||
>  		 (path && would_convert_to_git(istate, path)))
>  		ret = index_core(istate, oid, fd, xsize_t(st->st_size),

This is an old thread, but I happened across this code while looking at
the function for something more recent. This cast seems quite
suspicious. What happens when st->st_size is larger than a size_t (e.g.,
a 4.1GB file on a 32-bit system)? We'll truncate and get a wrong answer.

I think this ought to be comparing in the off_t space instead. Which I
_think_ was happening before the cast due to C's integer promotion rules
(even though off_t is signed, because it's larger than size_t on a
32-bit system and can hold all of size_t's values, then size_t is
promoted).

It's sort of academic, since both sides of the conditional end up
calling xsize_t(), which will then bail. But it _could_ handle off_t
correctly in the "else" clause here, which tries to stream the content.

More interesting to me (and why I responded to this old patch series) is
that inserting a cast to silence -Wsign-compare actually broke the
comparison in this case.

-Peff

  reply	other threads:[~2026-04-05  6:56 UTC|newest]

Thread overview: 60+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-07-09 11:17 [PATCH 00/19] object-file: get rid of `the_repository` Patrick Steinhardt
2025-07-09 11:17 ` [PATCH 01/19] object-file: fix -Wsign-compare warnings Patrick Steinhardt
2025-07-09 11:17 ` [PATCH 02/19] object-file: stop using `the_hash_algo` Patrick Steinhardt
2025-07-11  9:52   ` Karthik Nayak
2025-07-09 11:17 ` [PATCH 03/19] object-file: get rid of `the_repository` in `has_loose_object()` Patrick Steinhardt
2025-07-09 11:17 ` [PATCH 04/19] object-file: inline `check_and_freshen()` functions Patrick Steinhardt
2025-07-09 11:17 ` [PATCH 05/19] object-file: get rid of `the_repository` when freshening objects Patrick Steinhardt
2025-07-11  9:59   ` Karthik Nayak
2025-07-09 11:17 ` [PATCH 06/19] object-file: get rid of `the_repository` in `loose_object_info()` Patrick Steinhardt
2025-07-09 11:17 ` [PATCH 07/19] object-file: get rid of `the_repository` in `finalize_object_file()` Patrick Steinhardt
2025-07-09 11:17 ` [PATCH 08/19] loose: write loose objects map via their source Patrick Steinhardt
2025-07-11 10:25   ` Karthik Nayak
2025-07-15 10:50     ` Patrick Steinhardt
2025-07-09 11:17 ` [PATCH 09/19] odb: introduce `odb_write_object()` Patrick Steinhardt
2025-07-10 18:39   ` Toon Claes
2025-07-15 10:50     ` Patrick Steinhardt
2025-07-09 11:17 ` [PATCH 10/19] object-file: get rid of `the_repository` when writing objects Patrick Steinhardt
2025-07-09 11:17 ` [PATCH 11/19] object-file: inline `for_each_loose_file_in_objdir_buf()` Patrick Steinhardt
2025-07-09 11:17 ` [PATCH 12/19] object-file: remove declaration for `for_each_file_in_obj_subdir()` Patrick Steinhardt
2025-07-09 11:17 ` [PATCH 13/19] object-file: get rid of `the_repository` in loose object iterators Patrick Steinhardt
2025-07-10 18:41   ` Toon Claes
2025-07-09 11:17 ` [PATCH 14/19] object-file: get rid of `the_repository` in `read_loose_object()` Patrick Steinhardt
2025-07-09 11:17 ` [PATCH 15/19] object-file: get rid of `the_repository` in `force_object_loose()` Patrick Steinhardt
2025-07-10 18:42   ` Toon Claes
2025-07-11 10:38   ` Karthik Nayak
2025-07-15 10:50     ` Patrick Steinhardt
2025-07-15 11:36       ` Toon Claes
2025-07-09 11:17 ` [PATCH 16/19] object-file: get rid of `the_repository` in index-related functions Patrick Steinhardt
2025-07-09 11:17 ` [PATCH 17/19] environment: move compression level into repo settings Patrick Steinhardt
2025-07-09 15:26   ` Phillip Wood
2025-07-11 18:55     ` Junio C Hamano
2025-07-15 10:50       ` Patrick Steinhardt
2025-07-15 11:27       ` Patrick Steinhardt
2025-07-15 15:51         ` Phillip Wood
2025-07-15 16:12           ` Patrick Steinhardt
2025-07-16 12:56             ` Patrick Steinhardt
2025-07-17 15:19               ` Phillip Wood
2025-07-17 15:56                 ` Junio C Hamano
2025-07-15 18:50           ` Junio C Hamano
2025-07-17  8:00           ` Ayush Chandekar
2025-07-09 11:17 ` [PATCH 18/19] environment: move object creation mode " Patrick Steinhardt
2025-07-09 11:17 ` [PATCH 19/19] object-file: drop USE_THE_REPOSITORY_VARIABLE Patrick Steinhardt
2025-07-17  4:56 ` [PATCH v2 00/16] object-file: get rid of `the_repository` Patrick Steinhardt
2025-07-17  4:56   ` [PATCH v2 01/16] object-file: fix -Wsign-compare warnings Patrick Steinhardt
2026-04-05  6:56     ` Jeff King [this message]
2025-07-17  4:56   ` [PATCH v2 02/16] object-file: stop using `the_hash_algo` Patrick Steinhardt
2025-07-17  4:56   ` [PATCH v2 03/16] object-file: get rid of `the_repository` in `has_loose_object()` Patrick Steinhardt
2025-07-17  4:56   ` [PATCH v2 04/16] object-file: inline `check_and_freshen()` functions Patrick Steinhardt
2025-07-17  4:56   ` [PATCH v2 05/16] object-file: get rid of `the_repository` when freshening objects Patrick Steinhardt
2025-07-17  4:56   ` [PATCH v2 06/16] object-file: get rid of `the_repository` in `loose_object_info()` Patrick Steinhardt
2025-07-17  4:56   ` [PATCH v2 07/16] object-file: get rid of `the_repository` in `finalize_object_file()` Patrick Steinhardt
2025-07-17  4:56   ` [PATCH v2 08/16] loose: write loose objects map via their source Patrick Steinhardt
2025-07-17  4:56   ` [PATCH v2 09/16] odb: introduce `odb_write_object()` Patrick Steinhardt
2025-07-17  4:56   ` [PATCH v2 10/16] object-file: get rid of `the_repository` when writing objects Patrick Steinhardt
2025-07-17  4:56   ` [PATCH v2 11/16] object-file: inline `for_each_loose_file_in_objdir_buf()` Patrick Steinhardt
2025-07-17  4:56   ` [PATCH v2 12/16] object-file: remove declaration for `for_each_file_in_obj_subdir()` Patrick Steinhardt
2025-07-17  4:56   ` [PATCH v2 13/16] object-file: get rid of `the_repository` in loose object iterators Patrick Steinhardt
2025-07-17  4:56   ` [PATCH v2 14/16] object-file: get rid of `the_repository` in `read_loose_object()` Patrick Steinhardt
2025-07-17  4:56   ` [PATCH v2 15/16] object-file: get rid of `the_repository` in `force_object_loose()` Patrick Steinhardt
2025-07-17  4:56   ` [PATCH v2 16/16] object-file: get rid of `the_repository` in index-related functions Patrick Steinhardt

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20260405065602.GA1481314@coredump.intra.peff.net \
    --to=peff@peff.net \
    --cc=git@vger.kernel.org \
    --cc=gitster@pobox.com \
    --cc=karthik.188@gmail.com \
    --cc=phillip.wood123@gmail.com \
    --cc=ps@pks.im \
    --cc=toon@iotcl.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox