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

Re: [PATCH 4/5] tree-walk: unroll get_mode since loop boundaries are well-known

From
APAntriksh Pany <antriksh.pany@gmail.com>
Date
Apr 5, 2011, 23:55 UTC
Message-ID
<BANLkTin9P-OdTQhPTwcvgvpDoBg7E+va5Q@mail.gmail.com>
In-Reply-To
<BANLkTi==M=N+Z3qcsYk+tHap8A1Y41QfLw@mail.gmail.com>
On Mon, Apr 4, 2011 at 10:33 PM, Dan McGee <dpmcgee@gmail.com> wrote:
Show 8 quoted lines
> ....
> Totally agree; I should have tried to do it this way in the first
> place. However, compiling the fixed-length 0 to 5 loop does not
> produce fully-unrolled assembly for me with CFLAGS="-march=native
> -mtune=native -O2 -pipe -g" on x86_64. I see two copies of the loop
> only, and even worse is the (lack of) performance (each is the mode of
> 3 runs). Compilers are stupid apparently.
> ....

Can you try -O3? Or an explicit '-funroll-loops'? gcc I think does not do aggressive speed optimizations at the cost of space when at O2.

Cheers Antriksh

Previous: Dan McGeeNext: Dan McGee
Message 17 of 30 in “diff_tree_sha1: skip diff_tree if old == new”
  1. 1/5 diff_tree_sha1: skip diff_tree if old == newDan McGee, Mar 31, 2011
  2. 2/5 tree-walk: drop unused parameter from match_dir_prefixDan McGee, Mar 31, 2011
  3. Dan McGeeAug 30, 2011
  4. 3/5 tree-walk: micro-optimization in tree_entry_interestingDan McGee, Mar 31, 2011
  5. Nguyen Thai Ngoc DuyApr 3, 2011
  6. Junio C HamanoApr 3, 2011
  7. Dan McGeeApr 5, 2011
  8. tree_entry_interesting: inline strncmp()Nguyễn Thái Ngọc Duy, Apr 4, 2011
  9. 4/5 tree-walk: unroll get_mode since loop boundaries are well-knownDan McGee, Mar 31, 2011
  10. Nguyen Thai Ngoc DuyApr 2, 2011
  11. Dan McGeeApr 2, 2011
  12. Nguyen Thai Ngoc DuyApr 3, 2011
  13. Erik Faye-LundApr 4, 2011
  14. Andreas EricssonApr 4, 2011
  15. Junio C HamanoApr 4, 2011
  16. Dan McGeeApr 5, 2011
  17. Antriksh PanyApr 5, 2011
  18. Dan McGeeApr 6, 2011
  19. 5/5 tree-walk: match_entry microoptimizationDan McGee, Mar 31, 2011
  20. Nguyen Thai Ngoc DuyApr 2, 2011
  21. Dan McGeeApr 2, 2011
  22. Nguyen Thai Ngoc DuyMar 31, 2011
  23. Dan McGeeMar 31, 2011
  24. Junio C HamanoApr 1, 2011
  25. Nguyen Thai Ngoc DuyMay 3, 2011
  26. Fwd: [PATCH 1/5] diff_tree_sha1: skip diff_tree if old == newDan McGee, Apr 2, 2011
  27. Dan McGeeAug 30, 2011
  28. Junio C HamanoAug 30, 2011
  29. 1/2 tree-walk: drop unused parameter from match_dir_prefixDan McGee, Sep 9, 2011
  30. 2/2 tree-walk: micro-optimization in tree_entry_interestingDan McGee, Sep 9, 2011

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.