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

Re: [PATCH v3 1/2] auto gc: don't write bitmaps for incremental repacks

From
Jeff King <peff@peff.net>
Date
Dec 24, 2016, 02:57 UTC
Message-ID
<20161224025730.vwz2k4af6z6piinl@sigill.intra.peff.net>
In-Reply-To
<1482541735-21346-2-git-send-email-dturner@twosigma.com>
On Fri, Dec 23, 2016 at 08:08:54PM -0500, David Turner wrote:
> +	git gc --auto 2>err &&
> +	test_i18ngrep ! "^warning:" err &&
> +	ls .git/objects/pack/ | grep -v bitmap | sort >post_packs &&

I'm not sure why you omit the bitmap here. It shouldn't change, right? In that case it should not be mentioned by comm, and not impact the counts you check later (and in the off chance that a new .bitmap is created, we'd want to know and fail the test, I would think).

> +	comm --output-delimiter , -1 -3 existing_packs post_packs >new &&
> +	comm --output-delimiter , -2 -3 existing_packs post_packs >del &&
I don't think --output-delimiter will be portable beyond GNU comm.
> +	test_line_count = 0 del && # No packs are deleted
> +	test_line_count = 2 new # There is one new pack and its .idx
This logic makes sense (and is much nicer than the previous round).
-Peff
Previous: David Turner
Message 4 of 4 in “pack bitmaps and incremental packing”
  1. 0/2 pack bitmaps and incremental packingDavid Turner, Dec 24, 2016
  2. 2/2 repack: die on incremental + write-bitmap-indexDavid Turner, Dec 24, 2016
  3. 1/2 auto gc: don't write bitmaps for incremental repacksDavid Turner, Dec 24, 2016
  4. Jeff KingDec 24, 2016

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.