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

Re: [PATCH v3 5/5] archive-tar: use internal gzip by default

From
René Scharfe <l.s.r@web.de>
Date
Jun 14, 2022, 20:04 UTC
Message-ID
<4b4da5b3-7898-e97b-af74-a6874c8cb7e2@web.de>
In-Reply-To
<nycvar.QRO.7.76.6.2206141802310.353@tvgsbejvaqbjf.bet>
Am 14.06.22 um 18:29 schrieb Johannes Schindelin:
Show 54 quoted lines
> Hi René,
>
> On Tue, 14 Jun 2022, René Scharfe wrote:
>
>> Am 14.06.22 um 13:27 schrieb Johannes Schindelin:
>>
>>> The solution is to move the heap variable back into a scope that matches
>>> the lifetime of the compression:
>>>
>>> -- snip --
>>> diff --git a/archive-tar.c b/archive-tar.c
>>> index 60669eb7b9c..3d77e0f7509 100644
>>> --- a/archive-tar.c
>>> +++ b/archive-tar.c
>>> @@ -460,17 +460,12 @@ static void tgz_write_block(const void *data)
>>>
>>>  static const char internal_gzip_command[] = "git archive gzip";
>>>
>>> -static void tgz_set_os(git_zstream *strm, int os)
>>> -{
>>> -#if ZLIB_VERNUM >= 0x1221
>>> -	struct gz_header_s gzhead = { .os = os };
>>> -	deflateSetHeader(&strm->z, &gzhead);
>>> -#endif
>>> -}
>>> -
>>>  static int write_tar_filter_archive(const struct archiver *ar,
>>>  				    struct archiver_args *args)
>>>  {
>>> +#if ZLIB_VERNUM >= 0x1221
>>> +	struct gz_header_s gzhead = { .os = 3 }; /* Unix, for reproducibility */
>>> +#endif
>>>  	struct strbuf cmd = STRBUF_INIT;
>>>  	struct child_process filter = CHILD_PROCESS_INIT;
>>>  	int r;
>>> @@ -481,7 +476,10 @@ static int write_tar_filter_archive(const struct archiver *ar,
>>>  	if (!strcmp(ar->filter_command, internal_gzip_command)) {
>>>  		write_block = tgz_write_block;
>>>  		git_deflate_init_gzip(&gzstream, args->compression_level);
>>> -		tgz_set_os(&gzstream, 3); /* Unix, for reproducibility */
>>> +#if ZLIB_VERNUM >= 0x1221
>>> +		if (deflateSetHeader(&gzstream.z, &gzhead) != Z_OK)
>>> +			BUG("deflateSetHeader() called too late");
>>> +#endif
>>>  		gzstream.next_out = outbuf;
>>>  		gzstream.avail_out = sizeof(outbuf);
>>>
>>> -- snap --
>>
>> Good find, thank you!  A shorter solution would be to make gzhead static.
>
> I should have said that I had considered this, but decided against it
> because it would introduce yet another issue: it would render the code
> needlessly un-multi-threadable. And that can be avoided _really_ easily.

archive-tar.c (and archive-zip.c) use other static variables, so a static gzhead won't break or block anything in this regard. There was no interest in running it in parallel threads so far AFAIK, and it's hard for me to imagine the usefulness of creating multiple .tgz files at the same time.

The doubled ZLIB_VERNUM is unsightly and I'm not sure the BUG check is useful -- I omitted error checking because there is no recurse for us if deflateSetHeader() doesn't work, and on ancient zlib versions we silently continue anyway.

But that's all just minor quibbling -- I'll include your changes in the next version, they look fine overall.

René
Previous: Johannes SchindelinNext: Junio C Hamano
Message 48 of 74 in “Avoid spawning gzip in git archive”
  1. 0/2 Avoid spawning gzip in git archiveJohannes Schindelin via GitGitGadget, Apr 12, 2019
  2. 1/2 archive: replace write_or_die() calls with write_block_or_die()Rohit Ashiwal via GitGitGadget, Apr 12, 2019
  3. Jeff KingApr 13, 2019
  4. Junio C HamanoApr 13, 2019
  5. Rohit AshiwalApr 14, 2019
  6. Johannes SchindelinApr 26, 2019
  7. Junio C HamanoApr 26, 2019
  8. Johannes SchindelinApr 29, 2019
  9. Jeff KingMay 1, 2019
  10. René ScharfeMay 2, 2019
  11. Junio C HamanoMay 5, 2019
  12. Jeff KingMay 6, 2019
  13. Rohit AshiwalApr 14, 2019
  14. Junio C HamanoApr 14, 2019
  15. Johannes SchindelinApr 26, 2019
  16. Jeff KingMay 1, 2019
  17. 2/2 archive: avoid spawning `gzip`Rohit Ashiwal via GitGitGadget, Apr 12, 2019
  18. Jeff KingApr 13, 2019
  19. René ScharfeApr 13, 2019
  20. Jeff KingApr 15, 2019
  21. Johannes SchindelinApr 26, 2019
  22. René ScharfeApr 27, 2019
  23. René ScharfeApr 27, 2019
  24. Johannes SchindelinApr 29, 2019
  25. René ScharfeMay 1, 2019
  26. Jeff KingMay 1, 2019
  27. René ScharfeJun 10, 2019
  28. Jeff KingJun 13, 2019
  29. brian m. carlsonApr 13, 2019
  30. Jeff KingApr 15, 2019
  31. Johannes SchindelinApr 26, 2019
  32. Ævar Arnfjörð BjarmasonMay 2, 2019
  33. Johannes SchindelinMay 3, 2019
  34. Jeff KingMay 3, 2019
  35. Johannes SchindelinApr 26, 2019
  36. 0/5 Avoid spawning gzip in git archiveRené Scharfe, Jun 12, 2022
  37. 1/5 archive: rename archiver data field to filter_commandRené Scharfe, Jun 12, 2022
  38. 2/5 archive-tar: factor out write_block()René Scharfe, Jun 12, 2022
  39. 3/5 archive-tar: add internal gzip implementationRené Scharfe, Jun 12, 2022
  40. Junio C HamanoJun 13, 2022
  41. 4/5 archive-tar: use OS_CODE 3 (Unix) for internal gzipRené Scharfe, Jun 12, 2022
  42. 5/5 archive-tar: use internal gzip by defaultRené Scharfe, Jun 12, 2022
  43. Junio C HamanoJun 13, 2022
  44. Johannes SchindelinJun 14, 2022
  45. René ScharfeJun 14, 2022
  46. René ScharfeJun 14, 2022
  47. Johannes SchindelinJun 14, 2022
  48. René ScharfeJun 14, 2022
  49. Junio C HamanoJun 15, 2022
  50. Johannes SchindelinJun 14, 2022
  51. René ScharfeJun 14, 2022
  52. Johannes SchindelinJun 30, 2022
  53. Johannes SchindelinJul 1, 2022
  54. Jeff KingJul 1, 2022
  55. Junio C HamanoJul 1, 2022
  56. 0/6 Avoid spawning gzip in git archiveRené Scharfe, Jun 15, 2022
  57. 1/6 archive: update format documentationRené Scharfe, Jun 15, 2022
  58. 2/6 archive: rename archiver data field to filter_commandRené Scharfe, Jun 15, 2022
  59. 3/6 archive-tar: factor out write_block()René Scharfe, Jun 15, 2022
  60. 4/6 archive-tar: add internal gzip implementationRené Scharfe, Jun 15, 2022
  61. Ævar Arnfjörð BjarmasonJun 15, 2022
  62. René ScharfeJun 16, 2022
  63. Ævar Arnfjörð BjarmasonJun 24, 2022
  64. René ScharfeJun 24, 2022
  65. 5/6 archive-tar: use OS_CODE 3 (Unix) for internal gzipRené Scharfe, Jun 15, 2022
  66. 6/6 archive-tar: use internal gzip by defaultRené Scharfe, Jun 15, 2022
  67. René ScharfeMay 2, 2019
  68. René ScharfeMay 2, 2019
  69. Johannes SchindelinMay 8, 2019
  70. Jeff KingMay 8, 2019
  71. Johannes SchindelinMay 9, 2019
  72. Jeff KingMay 9, 2019
  73. René ScharfeMay 10, 2019
  74. Jeff KingMay 10, 2019

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.