public inbox for git@vger.kernel.org 
 help / color / mirror / Atom feed
From: Derrick Stolee <stolee@gmail•com>
To: Taylor Blau <me@ttaylorr•com>, Junio C Hamano <gitster@pobox•com>
Cc: Derrick Stolee via GitGitGadget <gitgitgadget@gmail•com>,
	git@vger•kernel.org
Subject: Re: [PATCH 3/5] midx-write: use cleanup when incremental midx fails
Date: Sat, 30 Aug 2025 10:44:22 -0400	[thread overview]
Message-ID: <eac886a7-0f50-40fa-b10e-92c68d60bac7@gmail.com> (raw)
In-Reply-To: <aLECkfDqFVgNA1xh@nand.local>

On 8/28/25 9:29 PM, Taylor Blau wrote:
> On Thu, Aug 28, 2025 at 01:51:18PM -0700, Junio C Hamano wrote:
>> "Derrick Stolee via GitGitGadget" <gitgitgadget@gmail•com> writes:
>>
>>> From: Derrick Stolee <stolee@gmail•com>
>>>
>>> The incremental mode of writing a multi-pack-index has a few extra
>>> conditions that could lead to failure, but these are currently
>>> short-ciruiting with 'return -1' instead of setting the method's
>>> 'result' variable and going to the cleanup tag.
>>>
>>> Replace these returns with gotos to avoid memory issues when exiting
>>> early due to error conditions.
>>>
>>> Unfortunately, these error conditions are difficult to reproduce with
>>> test cases, which is perhaps one reason why the memory loss was not
>>> caught by existing test cases in memory tracking modes.
>>>
>>> Signed-off-by: Derrick Stolee <stolee@gmail•com>
>>> ---
>>>   midx-write.c | 18 ++++++++++++------
>>>   1 file changed, 12 insertions(+), 6 deletions(-)
>>
>> Good thinking, but may I suggest us to go one more step to adopt
>> even better convention if we were to do this?
>>
>> Pessimistically initialize the "result" to -1 and let many "goto
>> cleanup" just jump there.  And have "result = 0" just before the
>> cleanup label where the success code path joins the final cleanup
>> part of the function.
>>
>> This is often the right way to make the flow easier to see, because
>> often the success code path is straight forward, and these error
>> conditions are what employ the "goto cleanup" from many places.  By
>> starting pessimistic, and declaring the success at the very end of
>> the straight-forward success case code path, all other flows to the
>> clean-up labels do not have to set the "ah I failed" flag.  It would
>> eliminate the need for patches like the previous step if the
>> original were following that pattern.
> 
> Alternatively replacing something like:
> 
>      error(...);
>      result = -1;
>      goto cleanup;
> 
> with just
> 
>      result = error(...);
>      goto cleanup;
> 
> would IMHO make the code easier to read, though I agree that nothing is
> forcing us to remember to assign result in the first place ;-). I am not
> sure the pessimistic initialization is better in all cases either, since
> we have to remember to place it before any "cleanup" label, and make
> sure that that does not regress.
> 
> So, I dunno. I'm OK with what is written here, and I think we could
> certainly have a separate discussion to perhaps have CodingGuidelines
> take a stronger stance here.

I'll look into adding this as a cleanup. This specific patch is
about adding the 'goto's where they were missing. I can make a new
patch that unifies the result initialization to -1 (and thus making
the method unified in returning -1 on error). There are many more
"result = 1" lines that need to change.

Thanks,
-Stolee

  reply	other threads:[~2025-08-30 14:44 UTC|newest]

Thread overview: 47+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-08-28 17:39 [PATCH 0/5] midx-write: fix segfault and do several cleanups Derrick Stolee via GitGitGadget
2025-08-28 17:39 ` [PATCH 1/5] midx-write: only load initialized packs Derrick Stolee via GitGitGadget
2025-08-28 20:19   ` Junio C Hamano
2025-08-29  1:20   ` Taylor Blau
2025-08-30 14:33     ` Derrick Stolee
2025-08-28 17:39 ` [PATCH 2/5] midx-write: put failing response value back Derrick Stolee via GitGitGadget
2025-08-28 20:45   ` Junio C Hamano
2025-08-29  1:26     ` Taylor Blau
2025-08-28 17:39 ` [PATCH 3/5] midx-write: use cleanup when incremental midx fails Derrick Stolee via GitGitGadget
2025-08-28 20:51   ` Junio C Hamano
2025-08-29  1:29     ` Taylor Blau
2025-08-30 14:44       ` Derrick Stolee [this message]
2025-08-28 17:39 ` [PATCH 4/5] midx-write: use uint32_t for preferred_pack_idx Derrick Stolee via GitGitGadget
2025-08-28 20:58   ` Junio C Hamano
2025-08-29  1:35   ` Taylor Blau
2025-08-28 17:39 ` [PATCH 5/5] midx-write: reenable signed comparison errors Derrick Stolee via GitGitGadget
2025-08-28 21:01   ` Junio C Hamano
2025-08-29  1:35     ` Taylor Blau
2025-08-29  1:36 ` [PATCH 0/5] midx-write: fix segfault and do several cleanups Taylor Blau
2025-08-30 21:23 ` [PATCH v2 0/6] " Derrick Stolee via GitGitGadget
2025-08-30 21:23   ` [PATCH v2 1/6] midx-write: only load initialized packs Derrick Stolee via GitGitGadget
2025-09-03 10:14     ` Patrick Steinhardt
2025-09-05 18:58       ` Derrick Stolee
2025-08-30 21:23   ` [PATCH v2 2/6] midx-write: put failing response value back Derrick Stolee via GitGitGadget
2025-09-03 10:15     ` Patrick Steinhardt
2025-09-05 19:03       ` Derrick Stolee
2025-08-30 21:23   ` [PATCH v2 3/6] midx-write: use cleanup when incremental midx fails Derrick Stolee via GitGitGadget
2025-09-03 10:15     ` Patrick Steinhardt
2025-08-30 21:23   ` [PATCH v2 4/6] midx-write: use uint32_t for preferred_pack_idx Derrick Stolee via GitGitGadget
2025-09-03 10:15     ` Patrick Steinhardt
2025-09-05 19:05       ` Derrick Stolee
2025-08-30 21:23   ` [PATCH v2 5/6] midx-write: reenable signed comparison errors Derrick Stolee via GitGitGadget
2025-09-03 10:15     ` Patrick Steinhardt
2025-08-30 21:23   ` [PATCH v2 6/6] midx-write: simplify error cases Derrick Stolee via GitGitGadget
2025-09-03 10:15     ` Patrick Steinhardt
2025-09-03 18:43       ` Junio C Hamano
2025-09-05 19:26   ` [PATCH v3 0/6] midx-write: fix segfault and do several cleanups Derrick Stolee via GitGitGadget
2025-09-05 19:26     ` [PATCH v3 1/6] midx-write: only load initialized packs Derrick Stolee via GitGitGadget
2025-09-05 19:26     ` [PATCH v3 2/6] midx-write: put failing response value back Derrick Stolee via GitGitGadget
2025-09-05 19:26     ` [PATCH v3 3/6] midx-write: use cleanup when incremental midx fails Derrick Stolee via GitGitGadget
2025-09-05 19:26     ` [PATCH v3 4/6] midx-write: use uint32_t for preferred_pack_idx Derrick Stolee via GitGitGadget
2025-09-05 19:26     ` [PATCH v3 5/6] midx-write: reenable signed comparison errors Derrick Stolee via GitGitGadget
2025-09-05 19:26     ` [PATCH v3 6/6] midx-write: simplify error cases Derrick Stolee via GitGitGadget
2025-09-05 19:38     ` [PATCH v3 0/6] midx-write: fix segfault and do several cleanups Junio C Hamano
2025-09-05 19:57       ` Derrick Stolee
2025-09-11 23:13         ` Taylor Blau
2025-09-11 23:44           ` Junio C Hamano

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=eac886a7-0f50-40fa-b10e-92c68d60bac7@gmail.com \
    --to=stolee@gmail$(echo .)com \
    --cc=git@vger$(echo .)kernel.org \
    --cc=gitgitgadget@gmail$(echo .)com \
    --cc=gitster@pobox$(echo .)com \
    --cc=me@ttaylorr$(echo .)com \
    /path/to/YOUR_REPLY

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