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

Re: obnoxious CLI complaints

From
Linus Torvalds <torvalds@linux-foundation.org>
Date
Sep 11, 2009, 14:47 UTC
Message-ID
<alpine.LFD.2.01.0909110744030.3654@localhost.localdomain>
In-Reply-To
<4AA97B61.6030301@lsrfire.ath.cx>
On Fri, 11 Sep 2009, René Scharfe wrote:
> 
> Using zlib directly avoids the overhead of a pipe and of buffering the
> output for blocked writes; surprisingly (to me), it isn't any faster.
In fact, it should be slower.

On SMP, you're quite likely better off using the pipe, and compressing on another CPU. Of course, it's usually the case that the compression is _so_ much slower than generating the tar-file (especially for the hot-cache case) that it doesn't matter or the pipe overhead is even a slowdown.

But especially if generating the tar-file has some delays in it (cold-cache object lookup, whatever), the "compress in separate process" is likely simply better, because you can compress while the other process is looking up data for the tar.

				Linus
Previous: René ScharfeNext: René Scharfe
Message 23 of 46 in “obnoxious CLI complaints”
  1. Brendan MillerSep 9, 2009
  2. Jakub NarebskiSep 9, 2009
  3. Wincent ColaiutaSep 9, 2009
  4. Jakub NarebskiSep 10, 2009
  5. Junio C HamanoSep 10, 2009
  6. René ScharfeSep 10, 2009
  7. Björn SteinbrinkSep 11, 2009
  8. John TapsellSep 10, 2009
  9. Sverre RabbelierSep 10, 2009
  10. Jakub NarebskiSep 10, 2009
  11. John TapsellSep 10, 2009
  12. Junio C HamanoSep 10, 2009
  13. demerphqSep 10, 2009
  14. Junio C HamanoSep 11, 2009
  15. John TapsellSep 11, 2009
  16. Junio C HamanoSep 11, 2009
  17. Brendan MillerSep 10, 2009
  18. Todd ZullingerSep 10, 2009
  19. Jakub NarebskiSep 10, 2009
  20. Eric SchaeferSep 10, 2009
  21. Sverre RabbelierSep 10, 2009
  22. René ScharfeSep 10, 2009
  23. Linus TorvaldsSep 11, 2009
  24. René ScharfeSep 11, 2009
  25. Linus TorvaldsSep 11, 2009
  26. Dmitry PotapovSep 12, 2009
  27. John TapsellSep 12, 2009
  28. Dmitry PotapovSep 12, 2009
  29. John TapsellSep 12, 2009
  30. A Large Angry SCMSep 12, 2009
  31. Dmitry PotapovSep 12, 2009
  32. John TapsellSep 12, 2009
  33. Junio C HamanoSep 13, 2009
  34. 1/2 git-archive: add '-o' as a alias for '--output'Dmitry Potapov, Sep 13, 2009
  35. 2/2 teach git-archive to auto detect the output formatDmitry Potapov, Sep 13, 2009
  36. Junio C HamanoSep 13, 2009
  37. 2/2 teach git-archive to auto detect the output formatDmitry Potapov, Sep 13, 2009
  38. Junio C HamanoSep 13, 2009
  39. Junio C HamanoSep 13, 2009
  40. 1/2 git-archive: add '-o' as a alias for '--output'Dmitry Potapov, Sep 13, 2009
  41. Brendan MillerSep 17, 2009
  42. Junio C HamanoSep 17, 2009
  43. Sverre RabbelierSep 9, 2009
  44. Pierre HabouzitSep 9, 2009
  45. Björn SteinbrinkSep 10, 2009
  46. Matthieu MoySep 10, 2009

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.