[PATCH] oneway_merge(): only lstat() when told to update worktree

[PATCH] oneway_merge(): only lstat() when told to update worktree

[Martin von Zweigbergk]

Although the subject line of 613f027 (read-tree -u one-way merge fix
to check out locally modified paths., 2006-05-15) mentions "read-tree
-u", it did not seem to check whether -u was in effect. Not checking
whether -u is in effect makes e.g. "read-tree --reset" lstat() the
worktree, even though the worktree stat should not matter for that
operation.

This speeds up e.g. "git reset" a little on the linux-2.6 repo (best
of five, warm cache):

        Before      After
real    0m0.288s    0m0.233s
user    0m0.190s    0m0.150s
sys     0m0.090s    0m0.080s
---

I am very unfamiliar with this part of git, so my attempt at a
motivation may be totally off.

I have another twenty-or-so patches to reset.c coming up that take the
timings down to around 90 ms, but this patch was quite unrelated to
that. Those other patches actually make this patch pointless for "git
reset" (it takes another path), but I hope this is still a good change
for other operations that use oneway_merge.

 unpack-trees.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/unpack-trees.c b/unpack-trees.c
index 6d96366..61acc5e 100644
--- a/unpack-trees.c
+++ b/unpack-trees.c
@@ -1834,7 +1834,7 @@ int oneway_merge(struct cache_entry **src, struct unpack_trees_options *o)
 
 	if (old && same(old, a)) {
 		int update = 0;
-		if (o->reset && !ce_uptodate(old) && !ce_skip_worktree(old)) {
+		if (o->reset && o->update && !ce_uptodate(old) && !ce_skip_worktree(old)) {
 			struct stat st;
 			if (lstat(old->name, &st) ||
 			    ie_match_stat(o->src_index, old, &st, CE_MATCH_IGNORE_VALID|CE_MATCH_IGNORE_SKIP_WORKTREE))
-- 
1.8.0.1.240.ge8a1f5a

Re: [PATCH] oneway_merge(): only lstat() when told to update worktree

[Junio C Hamano]

Martin von Zweigbergk <martinvonz@gmail.com> writes:

> Although the subject line of 613f027 (read-tree -u one-way merge fix
> to check out locally modified paths., 2006-05-15) mentions "read-tree
> -u", it did not seem to check whether -u was in effect. Not checking
> whether -u is in effect makes e.g. "read-tree --reset" lstat() the
> worktree, even though the worktree stat should not matter for that
> operation.
>
> This speeds up e.g. "git reset" a little on the linux-2.6 repo (best
> of five, warm cache):
>
>         Before      After
> real    0m0.288s    0m0.233s
> user    0m0.190s    0m0.150s
> sys     0m0.090s    0m0.080s
> ---

Sign-off?

I briefly discussed this with Martin in person and came to the same
conclusion. To me this looks like an obvious performance fix, but an
independent code audit catches our mistakes is of course welcomed.

Thanks.

> I am very unfamiliar with this part of git, so my attempt at a
> motivation may be totally off.
>
> I have another twenty-or-so patches to reset.c coming up that take the
> timings down to around 90 ms, but this patch was quite unrelated to
> that. Those other patches actually make this patch pointless for "git
> reset" (it takes another path), but I hope this is still a good change
> for other operations that use oneway_merge.
>
>  unpack-trees.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/unpack-trees.c b/unpack-trees.c
> index 6d96366..61acc5e 100644
> --- a/unpack-trees.c
> +++ b/unpack-trees.c
> @@ -1834,7 +1834,7 @@ int oneway_merge(struct cache_entry **src, struct unpack_trees_options *o)
>  
>  	if (old && same(old, a)) {
>  		int update = 0;
> -		if (o->reset && !ce_uptodate(old) && !ce_skip_worktree(old)) {
> +		if (o->reset && o->update && !ce_uptodate(old) && !ce_skip_worktree(old)) {
>  			struct stat st;
>  			if (lstat(old->name, &st) ||
>  			    ie_match_stat(o->src_index, old, &st, CE_MATCH_IGNORE_VALID|CE_MATCH_IGNORE_SKIP_WORKTREE))

[PATCH v2] oneway_merge(): only lstat() when told to update worktree

[Martin von Zweigbergk]

Although the subject line of 613f027 (read-tree -u one-way merge fix
to check out locally modified paths., 2006-05-15) mentions "read-tree
-u", it did not seem to check whether -u was in effect. Not checking
whether -u is in effect makes e.g. "read-tree --reset" lstat() the
worktree, even though the worktree stat should not matter for that
operation.

This speeds up e.g. "git reset" a little on the linux-2.6 repo (best
of five, warm cache):

        Before      After
real    0m0.288s    0m0.233s
user    0m0.190s    0m0.150s
sys     0m0.090s    0m0.080s

Signed-off-by: Martin von Zweigbergk <martinvonz@gmail.com>
---
 unpack-trees.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/unpack-trees.c b/unpack-trees.c
index 6d96366..61acc5e 100644
--- a/unpack-trees.c
+++ b/unpack-trees.c
@@ -1834,7 +1834,7 @@ int oneway_merge(struct cache_entry **src, struct unpack_trees_options *o)
 
 	if (old && same(old, a)) {
 		int update = 0;
-		if (o->reset && !ce_uptodate(old) && !ce_skip_worktree(old)) {
+		if (o->reset && o->update && !ce_uptodate(old) && !ce_skip_worktree(old)) {
 			struct stat st;
 			if (lstat(old->name, &st) ||
 			    ie_match_stat(o->src_index, old, &st, CE_MATCH_IGNORE_VALID|CE_MATCH_IGNORE_SKIP_WORKTREE))
-- 
1.8.0.1.240.ge8a1f5a