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

Re: [PATCH] index-pack, unpack-objects: increase input buffer from 4 KiB to 128 KiB

From
Junio C Hamano <gitster@pobox.com>
Date
Apr 25, 2026, 10:21 UTC
Message-ID
<xmqqldeb9w8e.fsf@gitster.g>
In-Reply-To
<pull.2282.git.git.1777058098756.gitgitgadget@gmail.com>
"Scott Bauersfeld via GitGitGadget" <gitgitgadget@gmail.com> writes:
Show 6 quoted lines
> From: Scott Bauersfeld <sbauersfeld@g.ucla.edu>
>
> Both index-pack and unpack-objects read pack data from stdin through
> a 4 KiB static buffer (input_buffer[4096]). On each fill(), consumed
> bytes are flushed to the output pack file via write_or_die(), so
> every write(2) moves at most 4 KiB.

Micronit. Output of unpack-objects obviously does not get flushed to "the output pack file".

Show 7 quoted lines
> On FUSE-backed filesystems every write(2) is a synchronous round
> trip through the FUSE protocol (userspace -> kernel -> userspace ->
> back), so the 4 KiB buffer turns a clone into many unnecessary tiny
> writes with noticeable latency overhead.
>
> Increase the buffer from 4 KiB to 128 KiB, matching the default
> already used by the hashfile layer in csum-file.c.
Quite sensible reasoning presented very nicely.
It may probably be a #leftoverbit but these three instances of (128
* 1024) may want to have a common symbolic constant, like
    #define DEFAULT_IOBUFFER_SIZE_IN_BYTES (128 * 1024)

in a bit more central header file. Especially for the one in csum-file.c where there is no symbolic constant used for that purpose.

Show 6 quoted lines
> Testing with strace on HTTPS clones of git/git (~296 MB pack, 5 runs
> per variant, isolated builds from the same v2.54.0 source) shows:
>
>   index-pack pack file writes: 72,465 -> 24,943 avg (66% reduction)
>   total write() syscalls:     310,192 -> 259,530 avg (17% reduction)
>   writes of exactly 4096 bytes: ~40,077 -> 0 (eliminated)

Hmph, I would have expected more like (1 - 4/128) ~ 97% reduction. The difference between that and 66% is coming from where? There are inherently short writes that do not utilize the new larger buffer beyond 4kB? If so, another number of interest might be the number of writes smaller than 4096 bytes, perhaps?

Previous: Scott Bauersfeld via GitGitGadgetNext: Derrick Stolee
Message 2 of 12 in “index-pack, unpack-objects: increase input buffer from 4 KiB to 128 KiB”
  1. index-pack, unpack-objects: increase input buffer from 4 KiB to 128 KiBScott Bauersfeld via GitGitGadget, Apr 24, 2026
  2. Junio C HamanoApr 25, 2026
  3. Derrick StoleeApr 27, 2026
  4. Junio C HamanoApr 28, 2026
  5. Jeff KingApr 28, 2026
  6. index-pack, unpack-objects: increase input buffer from 4 KiB to 128 KiBScott Bauersfeld via GitGitGadget, Apr 27, 2026
  7. Derrick StoleeApr 27, 2026
  8. index-pack, unpack-objects: increase input buffer from 4 KiB to 128 KiBScott Bauersfeld via GitGitGadget, Apr 27, 2026
  9. Derrick StoleeApr 27, 2026
  10. Junio C HamanoApr 28, 2026
  11. index-pack, unpack-objects: increase input buffer from 4 KiB to 128 KiBScott Bauersfeld via GitGitGadget, Apr 28, 2026
  12. Junio C HamanoMay 12, 2026

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.