On Mon, Apr 09 2012, Vladimir.Marek at oracle.com wrote:
> From: Vladimir Marek
>
> The inspiration was taken from similar issue in mutt:
> http://does-not-exist.org/mail-archives/mutt-dev/msg11290.html
>
> Signed-off-by: Vladimir Marek
> ---
Code looks pretty good, but 2 issues
1) Commit
Correct me if I'm mistaken, but d_name will only be a basename, so
your calls to stat will fail for files that are not in the current
directory. I think in all of the situations you had to call stat, we
already construct the absolute path of the file (sometimes a little
later in the code, but it
On Mon, Apr 09 2012, vladimir.ma...@oracle.com wrote:
From: Vladimir Marek vlma...@volny.cz
The inspiration was taken from similar issue in mutt:
http://does-not-exist.org/mail-archives/mutt-dev/msg11290.html
Signed-off-by: Vladimir Marek vlma...@volny.cz
---
Code looks pretty good, but 2
Correct me if I'm mistaken, but d_name will only be a basename, so
your calls to stat will fail for files that are not in the current
directory. I think in all of the situations you had to call stat, we
already construct the absolute path of the file (sometimes a little
later in the code, but it
From: Vladimir Marek
The inspiration was taken from similar issue in mutt:
http://does-not-exist.org/mail-archives/mutt-dev/msg11290.html
Signed-off-by: Vladimir Marek
---
notmuch-new.c | 28
1 files changed, 28 insertions(+), 0 deletions(-)
> Notmuch new is possibly one of the most performance critical bits for
> people with, uh, much mail. The performance impact of the new syscalls
> should be measured. (Can't do this myself atm.)
Fair enough. Is there some performance test suite? Another way would be
to make this compile time
On Apr 9, 2012 1:18 PM, wrote:
>
> From: Vladimir Marek
>
> The inspiration was taken from similar issue in mutt:
> http://does-not-exist.org/mail-archives/mutt-dev/msg11290.html
>
> Signed-off-by: Vladimir Marek
> ---
> notmuch-new.c | 19 +--
> 1 files changed, 13
From: Vladimir Marek
The inspiration was taken from similar issue in mutt:
http://does-not-exist.org/mail-archives/mutt-dev/msg11290.html
Signed-off-by: Vladimir Marek
---
notmuch-new.c | 19 +--
1 files changed, 13 insertions(+), 6 deletions(-)
diff --git
On Mon, Apr 9, 2012 at 09:12, Vladimir Marek
wrote:
> Fair enough. Is there some performance test suite? Another way would be
> to make this compile time option set by configure. Used only when the
> system in question does not have dirent->d_type member.
I like the idea of making it
From: Vladimir Marek vlma...@volny.cz
The inspiration was taken from similar issue in mutt:
http://does-not-exist.org/mail-archives/mutt-dev/msg11290.html
Signed-off-by: Vladimir Marek vlma...@volny.cz
---
notmuch-new.c | 19 +--
1 files changed, 13 insertions(+), 6
On Apr 9, 2012 1:18 PM, vladimir.ma...@oracle.com wrote:
From: Vladimir Marek vlma...@volny.cz
The inspiration was taken from similar issue in mutt:
http://does-not-exist.org/mail-archives/mutt-dev/msg11290.html
Signed-off-by: Vladimir Marek vlma...@volny.cz
---
notmuch-new.c | 19
Notmuch new is possibly one of the most performance critical bits for
people with, uh, much mail. The performance impact of the new syscalls
should be measured. (Can't do this myself atm.)
Fair enough. Is there some performance test suite? Another way would be
to make this compile time option
From: Vladimir Marek vlma...@volny.cz
The inspiration was taken from similar issue in mutt:
http://does-not-exist.org/mail-archives/mutt-dev/msg11290.html
Signed-off-by: Vladimir Marek vlma...@volny.cz
---
notmuch-new.c | 28
1 files changed, 28 insertions(+), 0
13 matches
Mail list logo