Re: [PATCH] ls -i: print consistent inode numbers also for mount points

2009-09-01 Thread Jim Meyering
Eric Blake wrote: > Have we reported this to the Linux developers, along with chapter and > verse of POSIX 2008 which shows they are non-compliant? I've just written to lkml: http://marc.info/?l=linux-kernel&m=125181054102075

Re: [PATCH] ls -i: print consistent inode numbers also for mount points

2009-09-01 Thread Eric Blake
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 According to Ondřej Vašík on 8/31/2009 7:07 AM: >> It's a shame to have to pessimize ls -i performance, even by this >> small amount on linux-based kernels, but correctness definitely >> trumps performance, here. > > Yep, it has some performance impac

Re: [PATCH] ls -i: print consistent inode numbers also for mount points

2009-08-31 Thread Jim Meyering
Jim Meyering wrote: > Following up on a long thread from a year ago, here's a patch > to fix the 3.5-year-old readdir-vs-mountpoint-inode bug in ls -i. ... > http://thread.gmane.org/gmane.comp.gnu.coreutils.bugs/14020 ... > It's a shame to have to pessimize ls -i performance, even by this > small a

Re: [PATCH] ls -i: print consistent inode numbers also for mount points

2009-08-31 Thread Jim Meyering
Pádraig Brady wrote: ... >> Yep, it has some performance impact... >> checked `time ./ls -i1R /home /dev /usr /var /lib >myinodes with >> approximate number of 31k dirs, 450k files on ext3 >> >> and results are >> old binary without the patch: >> real2m5.631s >> user0m3.012s >> sys 0m4.

Re: [PATCH] ls -i: print consistent inode numbers also for mount points

2009-08-31 Thread Pádraig Brady
Ondřej Vašík wrote: > Pádraig Brady wrote: >> So there was a lot more disk access with the new binary. >> but the NEWS suggests that should only be the case for >> "systems with dysfunctional readdir". >> >> What was your system? > > Quite ancient system... I checked this on my old Fedora Core 6..

Re: [PATCH] ls -i: print consistent inode numbers also for mount points

2009-08-31 Thread Ondřej Vašík
Pádraig Brady wrote: > So there was a lot more disk access with the new binary. > but the NEWS suggests that should only be the case for > "systems with dysfunctional readdir". > > What was your system? Quite ancient system... I checked this on my old Fedora Core 6... Anyway could check with some

Re: [PATCH] ls -i: print consistent inode numbers also for mount points

2009-08-31 Thread Pádraig Brady
Ondřej Vašík wrote: > Jim Meyering wrote: >> Following up on a long thread from a year ago, here's a patch >> to fix the 3.5-year-old readdir-vs-mountpoint-inode bug in ls -i. > > Checked on my system and works fine... > >> It's a shame to have to pessimize ls -i performance, even by this >> smal

Re: [PATCH] ls -i: print consistent inode numbers also for mount points

2009-08-31 Thread Jim Meyering
Ondřej Vašík wrote: > Jim Meyering wrote: >> Following up on a long thread from a year ago, here's a patch >> to fix the 3.5-year-old readdir-vs-mountpoint-inode bug in ls -i. > > Checked on my system and works fine... > >> It's a shame to have to pessimize ls -i performance, even by this >> small

Re: [PATCH] ls -i: print consistent inode numbers also for mount points

2009-08-31 Thread Ondřej Vašík
Jim Meyering wrote: > Following up on a long thread from a year ago, here's a patch > to fix the 3.5-year-old readdir-vs-mountpoint-inode bug in ls -i. Checked on my system and works fine... > It's a shame to have to pessimize ls -i performance, even by this > small amount on linux-based kernels,

[PATCH] ls -i: print consistent inode numbers also for mount points

2009-08-30 Thread Jim Meyering
t-7.6. >From 3af748aa25193e8a5a8fe520cd967cfbc4d71cb8 Mon Sep 17 00:00:00 2001 From: Jim Meyering Date: Wed, 2 Jul 2008 18:01:43 +0200 Subject: [PATCH] ls -i: print consistent inode numbers also for mount points On most unix- and linux-based kernels, ls -i DIR_CONTAINING_MOUNT_POINT would print the wrong in