debian/patches/0015-is_ntfs_dotgit-speed-it-up.diff

   1 From b281c877c67632aa62fbcaedd07f7981739970fa Mon Sep 17 00:00:00 2001
   2 From: Johannes Schindelin <johannes.schindelin@gmx.de>
   3 Date: Fri, 6 Sep 2019 21:09:35 +0200
   4 Subject: is_ntfs_dotgit(): speed it up
   5
   6 Previously, this function was written without focusing on speed,
   7 intending to make reviewing the code as easy as possible, to avoid any
   8 bugs in this critical code.
   9
  10 Turns out: we can do much better on both accounts. With this patch, we
  11 make it as fast as this developer can make it go:
  12
  13 - We avoid the call to `is_dir_sep()` and make all the character
  14   comparisons explicit.
  15
  16 - We avoid the cost of calling `strncasecmp()` and unroll the test for
  17   `.git` and `git~1`, not even using `tolower()` because it is faster to
  18   compare against two constant values.
  19
  20 - We look for `.git` and `.git~1` first thing, and return early if not
  21   found.
  22
  23 - We also avoid calling a separate function for detecting chains of
  24   spaces and periods.
  25
  26 Each of these improvements has a noticeable impact on the speed of
  27 `is_ntfs_dotgit()`.
  28
  29 Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
  30 (cherry picked from commit 3a85dc7d534fc2d410ddc0c771c963b20d1b4857)
  31 Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
  32 ---
  33  path.c | 55 ++++++++++++++++++++++++++++++-------------------------
  34  1 file changed, 30 insertions(+), 25 deletions(-)
  35
  36 diff --git a/path.c b/path.c
  37 index 07d20285e6..e42bba6686 100644
  38 --- a/path.c
  39 +++ b/path.c
  40 @@ -1316,20 +1316,6 @@ int daemon_avoid_alias(const char *p)
  41         }
  42  }
  43
  44 -static int only_spaces_and_periods(const char *path, size_t len, size_t skip)
  45 -{
  46 -       if (len < skip)
  47 -               return 0;
  48 -       len -= skip;
  49 -       path += skip;
  50 -       while (len-- > 0) {
  51 -               char c = *(path++);
  52 -               if (c != ' ' && c != '.')
  53 -                       return 0;
  54 -       }
  55 -       return 1;
  56 -}
  57 -
  58  /*
  59   * On NTFS, we need to be careful to disallow certain synonyms of the `.git/`
  60   * directory:
  61 @@ -1369,19 +1355,38 @@ static int only_spaces_and_periods(const char *path, size_t len, size_t skip)
  62   */
  63  int is_ntfs_dotgit(const char *name)
  64  {
  65 -       size_t len;
  66 +       char c;
  67
  68 -       for (len = 0; ; len++)
  69 -               if (!name[len] || name[len] == '\\' || is_dir_sep(name[len]) ||
  70 -                   name[len] == ':') {
  71 -                       if (only_spaces_and_periods(name, len, 4) &&
  72 -                                       !strncasecmp(name, ".git", 4))
  73 -                               return 1;
  74 -                       if (only_spaces_and_periods(name, len, 5) &&
  75 -                                       !strncasecmp(name, "git~1", 5))
  76 -                               return 1;
  77 +       /*
  78 +        * Note that when we don't find `.git` or `git~1` we end up with `name`
  79 +        * advanced partway through the string. That's okay, though, as we
  80 +        * return immediately in those cases, without looking at `name` any
  81 +        * further.
  82 +        */
  83 +       c = *(name++);
  84 +       if (c == '.') {
  85 +               /* .git */
  86 +               if (((c = *(name++)) != 'g' && c != 'G') ||
  87 +                   ((c = *(name++)) != 'i' && c != 'I') ||
  88 +                   ((c = *(name++)) != 't' && c != 'T'))
  89                         return 0;
  90 -               }
  91 +       } else if (c == 'g' || c == 'G') {
  92 +               /* git ~1 */
  93 +               if (((c = *(name++)) != 'i' && c != 'I') ||
  94 +                   ((c = *(name++)) != 't' && c != 'T') ||
  95 +                   *(name++) != '~' ||
  96 +                   *(name++) != '1')
  97 +                       return 0;
  98 +       } else
  99 +               return 0;
 100 +
 101 +       for (;;) {
 102 +               c = *(name++);
 103 +               if (!c || c == '\\' || c == '/' || c == ':')
 104 +                       return 1;
 105 +               if (c != '.' && c != ' ')
 106 +                       return 0;
 107 +       }
 108  }
 109
 110  static int is_ntfs_dot_generic(const char *name,
 111 --
 112 2.24.0.393.g34dc348eaf
 113