* [PATCH v2] Make git prune remove temporary packs that look like write failures
@ 2008-02-06 23:48 David Steven Tweed
2008-02-07 1:25 ` Johannes Schindelin
0 siblings, 1 reply; 2+ messages in thread
From: David Steven Tweed @ 2008-02-06 23:48 UTC (permalink / raw)
To: git; +Cc: nico, Johannes.Schindelin, casey
Write errors when repacking (eg, due to out-of-space conditions)
can leave temporary packs (and possibly other files beginning
with "tmp_") lying around which no existing
codepath removes and which aren't obvious to the casual user.
These can also be multi-megabyte files wasting noticeable space.
Unfortunately there's no way to definitely tell in builtin-prune
that a tmp_ file is not being used by a concurrent process.
However, it is documented that pruning should only be done
on a quiet repository and --expire is honoured (using code
from Johannes Schindelin). The names of removed files are printed.
Signed-off-by: David Tweed (david.tweed@gmail.com)
---
This version incorporates all the agreed with changes from the
previous version, and since I was modifying the relevant
context lines it includes Johannes' code honouring --expire
(but not the test he wrote which is presumably best sent as a separate
patch from him).
Hopefully this one won't get whitespace munged.
builtin-prune.c | 39 +++++++++++++++++++++++++++++++++++++++
1 files changed, 39 insertions(+), 0 deletions(-)
diff --git a/builtin-prune.c b/builtin-prune.c
index b5e7684..f25ddb6 100644
--- a/builtin-prune.c
+++ b/builtin-prune.c
@@ -83,6 +83,44 @@ static void prune_object_dir(const char *path)
}
}
+/*
+ * Write errors (particularly out of space) can result in
+ * failed temporary packs (and more rarely indexes and other
+ * files begining with "tmp_") accumulating in the
+ * object directory.
+ */
+static void remove_temporary_files(void)
+{
+ DIR *dir;
+ struct dirent *de;
+ char* dirname=get_object_directory();
+
+ dir = opendir(dirname);
+ if (!dir) {
+ fprintf(stderr, "Unable to open object directory %s\n",
+ dirname);
+ return;
+ }
+ while ((de = readdir(dir)) != NULL) {
+ if (!prefixcmp(de->d_name, "tmp_")) {
+ char name[PATH_MAX];
+ int c = snprintf(name, PATH_MAX, "%s/%s",
+ dirname, de->d_name);
+ if (c < 0 || c >= PATH_MAX)
+ continue;
+ if (expire) {
+ struct stat st;
+ if (stat(name, &st) != 0 || st.st_mtime >= expire)
+ continue;
+ }
+ printf("Removing stale temporary file %s\n", name);
+ if (!show_only)
+ unlink(name);
+ }
+ }
+ closedir(dir);
+}
+
int cmd_prune(int argc, const char **argv, const char *prefix)
{
int i;
@@ -115,5 +153,6 @@ int cmd_prune(int argc, const char **argv, const char *prefix)
sync();
prune_packed_objects(show_only);
+ remove_temporary_files();
return 0;
}
^ permalink raw reply related [flat|nested] 2+ messages in thread
* Re: [PATCH v2] Make git prune remove temporary packs that look like write failures
2008-02-06 23:48 [PATCH v2] Make git prune remove temporary packs that look like write failures David Steven Tweed
@ 2008-02-07 1:25 ` Johannes Schindelin
0 siblings, 0 replies; 2+ messages in thread
From: Johannes Schindelin @ 2008-02-07 1:25 UTC (permalink / raw)
To: David Steven Tweed; +Cc: git, nico, casey
Hi,
On Wed, 6 Feb 2008, David Steven Tweed wrote:
> This version incorporates all the agreed with changes from the previous
> version, and since I was modifying the relevant context lines it
> includes Johannes' code honouring --expire (but not the test he wrote
> which is presumably best sent as a separate patch from him).
No, I don't want the tests as a separate patch. They are intrinsic to
your patch, since they show that your patch does what it's supposed to do
(at least in an empirical way).
Ciao,
Dscho
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2008-02-07 1:26 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2008-02-06 23:48 [PATCH v2] Make git prune remove temporary packs that look like write failures David Steven Tweed
2008-02-07 1:25 ` Johannes Schindelin
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox