Re: [bug #24342] -inum predicate shoud use dirent.d_ino instead of stat.st_ino

2009-03-08 Thread George Spelvin
James Youngman wrote: > [ adding nug-gnulib to CC since we're discussinf its fts implementation ] > > On Sun, Mar 8, 2009 at 2:50 PM, George Spelvin wrote: >> James Youngman wrote: >>> The updated patches are what I actually pushed (into 4.5.x). >> >> Ah, thank you. The posted patches wouldn't

Re: [PATCH] Switch to using a merge driver for the ChangeLog file.

2009-03-08 Thread Eric Blake
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 According to James Youngman on 3/8/2009 7:43 AM: > * .gitattributes: new file You went with the name merge.cl-merge.driver, while gnulib, coreutils, autoconf, libtool, and m4 all used merge.merge-changelog.driver. I think it would be nice to go with

Re: [Findutils-patches] [PATCH] Optimise away calls to stat if all we need is the inode number (bug #24342).

2009-03-08 Thread Jim Meyering
James Youngman wrote: > On Sun, Mar 8, 2009 at 4:35 PM, Jim Meyering wrote: >> James Youngman wrote: >> ... >>> +#24342: -inum predicate shoud use dirent.d_ino instead of stat.st_ino >>> +(this is a performance bug). >> >> You might want to mention that dirent.d_ino values are unreliable on some >

Re: [Findutils-patches] [PATCH] Optimise away calls to stat if all we need is the inode number (bug #24342).

2009-03-08 Thread James Youngman
On Sun, Mar 8, 2009 at 4:35 PM, Jim Meyering wrote: > James Youngman wrote: > ... >> +#24342: -inum predicate shoud use dirent.d_ino instead of stat.st_ino >> +(this is a performance bug). > > You might want to mention that dirent.d_ino values are unreliable on some > systems (Linux, Solaris, *BSD

Re: [Findutils-patches] [PATCH] Optimise away calls to stat if all we need is the inode number (bug #24342).

2009-03-08 Thread Jim Meyering
James Youngman wrote: ... > +#24342: -inum predicate shoud use dirent.d_ino instead of stat.st_ino > +(this is a performance bug). You might want to mention that dirent.d_ino values are unreliable on some systems (Linux, Solaris, *BSD get it wrong, but Cygwin gets it right). The problem is with mo

Re: [bug #24342] -inum predicate shoud use dirent.d_ino instead of stat.st_ino

2009-03-08 Thread James Youngman
[ adding nug-gnulib to CC since we're discussinf its fts implementation ] On Sun, Mar 8, 2009 at 2:50 PM, George Spelvin wrote: > James Youngman wrote: >> The updated patches are what I actually pushed (into 4.5.x). > > Ah, thank you.  The posted patches wouldn't apply to savannah git HEAD > due

Re: [bug #24342] -inum predicate shoud use dirent.d_ino instead of stat.st_ino

2009-03-08 Thread George Spelvin
James Youngman wrote: > The updated patches are what I actually pushed (into 4.5.x). Ah, thank you. The posted patches wouldn't apply to savannah git HEAD due to a conflict in the ChangeLog. (Easy enough to work around, but worrisome.) The (tested) simplified version is, of course, + boolean

Re: [Findutils-patches] [PATCH] Delete the INSTALL file; rely on Autoconf providing it.

2009-03-08 Thread James Youngman
On Fri, Mar 6, 2009 at 10:27 AM, James Youngman wrote: > * INSTALL: Delete; this file can be copied/generated by Autoconf, > so do that rather than keeping it under source control. > * .gitignore: Ignore INSTALL. Applied and pushed. James.

Re: [PATCH] Switch to using a merge driver for the ChangeLog file.

2009-03-08 Thread James Youngman
On Sun, Mar 8, 2009 at 1:43 PM, James Youngman wrote: > * import-gnulib.sh (check_merge_driver): New function; verifies > that we have a merge driver installed for the ChangeLog file (if > we're on the master or release branch, the check is skipped). > Also check that the .gitattributes file exist

[PATCH] Switch to using a merge driver for the ChangeLog file.

2009-03-08 Thread James Youngman
* import-gnulib.sh (check_merge_driver): New function; verifies that we have a merge driver installed for the ChangeLog file (if we're on the master or release branch, the check is skipped). Also check that the .gitattributes file exists. * README-hacking: Point out that using a merge driver for Ch

[bug #24342] -inum predicate shoud use dirent.d_ino instead of stat.st_ino

2009-03-08 Thread James Youngman
Update of bug #24342 (project findutils): Status:None => Fixed ___ Follow-up Comment #2: The updated patches are what I actually pushed (into 4.5.x). (file #17618, file #17619)

Re: [bug #24342] -inum predicate shoud use dirent.d_ino instead of stat.st_ino

2009-03-08 Thread George Spelvin
(Resend because I can't spell "bug-findutils".) Thanks for the fix! Sorry it fell off the bottom of my to-do pile. If I might suggest, the code here does not express the logic very clearly: diff --git a/find/util.c b/find/util.c index 2570682..bbc8436 100644 --- a/find/util.c +++ b/find/util.c