Bug#683768: linux-image-3.2.0-3-amd64: hibernate 3x slower after upgrade to linux-image-3.2.0-3-amd64
It gets stranger still: when I compile a kernel after 'make localmodconfig' to streamline it, then the hibernate time sometimes changes. This is true for 3.2.23-1: it goes from 31s. to 10s. It is not true for 3.2.21-3: it stays at 31s. Neither is it true for 3.5.0 from experimental: it was 30s. and streamlined only went to 20s. Hugo
Bug#683768: linux-image-3.2.0-3-amd64: hibernate 3x slower after upgrade to linux-image-3.2.0-3-amd64
On Sun, Aug 5, 2012 at 6:36 PM, Ben Hutchings b...@decadent.org.uk wrote: On Sun, 2012-08-05 at 17:50 -0500, hugo vanwoerkom wrote: [...] Yes, 3.2.19-1 still takes 12s. But as to problems disappearing/appearing I could have sworn that 3.2.21-3 took 12s. last night but now it takes 35s. I think the problem comes from the kernel, because that is the only thing that I am varying and I have been hibernating this box for years with up to now predictable speeds. Don't use wicd though. If the change was made between 3.2.19 and 3.2.21 then it might be due to: commit d006ab31cd818f5e4dda2453fd09767063f49933 Author: Michal Hocko mho...@suse.cz Date: Tue May 29 15:06:45 2012 -0700 mm: consider all swapped back pages in used-once logic So you could test with the reverse of that patch (attached), following the instructions at http://kernel-handbook.alioth.debian.org/ch-common-tasks.html#s4.2.2 I tested all the snapshot kernels again and the change occurred with 3.2.21-1. That is the first one to have a hibernate of 35s. The one previous to that 3.2.20-1 is fine with 10s. These results are consistent. It could not be the mm vmscan patch because both 3.2.20-1 and 3.2.21-1 have that set the same way: if (referenced_ptes) { if (PageSwapBacked(page)) return PAGEREF_ACTIVATE; Hugo
Bug#683768: linux-image-3.2.0-3-amd64: hibernate 3x slower after upgrade to linux-image-3.2.0-3-amd64
Slow hibernate occurs for the first time in 3.2.23-1. None of the kernels previous to that one have the problem. 3.5-1~experimental.1 still has the problem. Thanks! Hugo
Bug#683768: linux-image-3.2.0-3-amd64: hibernate 3x slower after upgrade to linux-image-3.2.0-3-amd64
On Sun, 2012-08-05 at 12:11 -0500, hugo vanwoerkom wrote: Slow hibernate occurs for the first time in 3.2.23-1. I really don't understand this, because I can't see any changes between 3.2.21-3 and 3.2.23-1 that could slow down disk I/O for hibernation. My suspicion is that this has something to do with the amount of memory that needs to be saved. But I also don't see why that would increase by any significant amount. None of the kernels previous to that one have the problem. 3.5-1~experimental.1 still has the problem. Please report this upstream at https://bugzilla.kernel.org under product 'Power Management', component 'Hibernation/Suspend'. Ben. -- Ben Hutchings Computers are not intelligent. They only think they are. signature.asc Description: This is a digitally signed message part
Bug#683768: linux-image-3.2.0-3-amd64: hibernate 3x slower after upgrade to linux-image-3.2.0-3-amd64
On Sun, Aug 5, 2012 at 2:47 PM, Ben Hutchings b...@decadent.org.uk wrote: On Sun, 2012-08-05 at 12:11 -0500, hugo vanwoerkom wrote: Slow hibernate occurs for the first time in 3.2.23-1. I really don't understand this, because I can't see any changes between 3.2.21-3 and 3.2.23-1 that could slow down disk I/O for hibernation. My suspicion is that this has something to do with the amount of memory that needs to be saved. But I also don't see why that would increase by any significant amount. None of the kernels previous to that one have the problem. 3.5-1~experimental.1 still has the problem. Please report this upstream at https://bugzilla.kernel.org under product 'Power Management', component 'Hibernation/Suspend'. Let me check some more Ben. I just checked 3.2.21-3 again and it took 40secs. to hibernate. Should be 12secs. I'll let you know for sure. Hugo
Bug#683768: linux-image-3.2.0-3-amd64: hibernate 3x slower after upgrade to linux-image-3.2.0-3-amd64
On 2012-08-05 15:27:47 -0500, hugo vanwoerkom wrote: Let me check some more Ben. I just checked 3.2.21-3 again and it took 40secs. to hibernate. Should be 12secs. I'll let you know for sure. Have you tried to downgrade to 3.2.19-1 to see if it still takes 12s? It seems that problems can strangely appear and disappear without any known reason. Are you sure that the problem comes from the kernel itself and not from one of the hooks (see /var/log/pm-suspend.log)? For me, the slow suspend (not always reproducible) comes from the wicd hook. -- Vincent Lefèvre vinc...@vinc17.net - Web: http://www.vinc17.net/ 100% accessible validated (X)HTML - Blog: http://www.vinc17.net/blog/ Work: CR INRIA - computer arithmetic / AriC project (LIP, ENS-Lyon) -- To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20120805211729.ga4...@xvii.vinc17.org
Bug#683768: linux-image-3.2.0-3-amd64: hibernate 3x slower after upgrade to linux-image-3.2.0-3-amd64
On Sun, Aug 5, 2012 at 4:17 PM, Vincent Lefevre vinc...@vinc17.net wrote: On 2012-08-05 15:27:47 -0500, hugo vanwoerkom wrote: Let me check some more Ben. I just checked 3.2.21-3 again and it took 40secs. to hibernate. Should be 12secs. I'll let you know for sure. Have you tried to downgrade to 3.2.19-1 to see if it still takes 12s? It seems that problems can strangely appear and disappear without any known reason. Are you sure that the problem comes from the kernel itself and not from one of the hooks (see /var/log/pm-suspend.log)? For me, the slow suspend (not always reproducible) comes from the wicd hook. Yes, 3.2.19-1 still takes 12s. But as to problems disappearing/appearing I could have sworn that 3.2.21-3 took 12s. last night but now it takes 35s. I think the problem comes from the kernel, because that is the only thing that I am varying and I have been hibernating this box for years with up to now predictable speeds. Don't use wicd though. Hugo
Bug#683768: linux-image-3.2.0-3-amd64: hibernate 3x slower after upgrade to linux-image-3.2.0-3-amd64
On Sun, 2012-08-05 at 17:50 -0500, hugo vanwoerkom wrote: [...] Yes, 3.2.19-1 still takes 12s. But as to problems disappearing/appearing I could have sworn that 3.2.21-3 took 12s. last night but now it takes 35s. I think the problem comes from the kernel, because that is the only thing that I am varying and I have been hibernating this box for years with up to now predictable speeds. Don't use wicd though. If the change was made between 3.2.19 and 3.2.21 then it might be due to: commit d006ab31cd818f5e4dda2453fd09767063f49933 Author: Michal Hocko mho...@suse.cz Date: Tue May 29 15:06:45 2012 -0700 mm: consider all swapped back pages in used-once logic So you could test with the reverse of that patch (attached), following the instructions at http://kernel-handbook.alioth.debian.org/ch-common-tasks.html#s4.2.2 Ben. -- Ben Hutchings Theory and practice are closer in theory than in practice. - John Levine, moderator of comp.compilers From f5487f80a6e4561b38855f4b5fd1fb94b203e755 Mon Sep 17 00:00:00 2001 From: Ben Hutchings b...@decadent.org.uk Date: Mon, 6 Aug 2012 00:30:00 +0100 Subject: [PATCH] Revert mm: consider all swapped back pages in used-once logic This reverts commit d006ab31cd818f5e4dda2453fd09767063f49933. Testing for bug #683768. --- mm/vmscan.c |2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/mm/vmscan.c b/mm/vmscan.c index 48febd7..79b3ef9 100644 --- a/mm/vmscan.c +++ b/mm/vmscan.c @@ -697,7 +697,7 @@ static enum page_references page_check_references(struct page *page, return PAGEREF_RECLAIM; if (referenced_ptes) { - if (PageSwapBacked(page)) + if (PageAnon(page)) return PAGEREF_ACTIVATE; /* * All mapped pages start out with page table signature.asc Description: This is a digitally signed message part
Bug#683768: linux-image-3.2.0-3-amd64: hibernate 3x slower after upgrade to linux-image-3.2.0-3-amd64
Unfortunately the swap.c patch makes no difference at all. :-( Hugo
Bug#683768: linux-image-3.2.0-3-amd64: hibernate 3x slower after upgrade to linux-image-3.2.0-3-amd64
On Sat, 2012-08-04 at 07:21 -0500, hugo vanwoerkom wrote: Unfortunately the swap.c patch makes no difference at all. :-( OK, so this a different problem. Since this regression occurred between 3.2.19-1 and 3.2.23-1, can you bisect through the intermediate versions available from http://snapshot.debian.org/package/linux/: 1. Test version 3.2.21-2 2. If it has this bug, test an earlier version, otherwise test a later version. 3. Repeat if necessary to identify the first buggy version. Also, please test 3.5-1~experimental.1 (from the experimental suite) to see whether this has been fixed upstream. Ben. -- Ben Hutchings Experience is directly proportional to the value of equipment destroyed. - Carolyn Scheppner signature.asc Description: This is a digitally signed message part
Bug#683768: linux-image-3.2.0-3-amd64: hibernate 3x slower after upgrade to linux-image-3.2.0-3-amd64
But it's a Debian problem. I compiled the stable 3.4.7 kernel from kernel.org and that does not have the problem: normal hibernate, takes 12secs. Hugo
Bug#683768: linux-image-3.2.0-3-amd64: hibernate 3x slower after upgrade to linux-image-3.2.0-3-amd64
On Fri, Aug 3, 2012 at 12:06:13 -0500, hugo vanwoerkom wrote: Package: src:linux Version: 3.2.23-1 Severity: normal Hibernate of about 400.000 pages takes normally 12secs. after the last upgrade of the kernel (wheezy or sid) it takes 40 secs. It is not the disk space because when I move that to a different device, the same thing happens. This might be a feature, but it is so noticable :-( Thanks! Is that reproducible without the nvidia driver? Cheers, Julien signature.asc Description: Digital signature
Bug#683768: linux-image-3.2.0-3-amd64: hibernate 3x slower after upgrade to linux-image-3.2.0-3-amd64
On Fri, Aug 3, 2012 at 12:36 PM, Julien Cristau jcris...@debian.org wrote: On Fri, Aug 3, 2012 at 12:06:13 -0500, hugo vanwoerkom wrote: Package: src:linux Version: 3.2.23-1 Severity: normal Hibernate of about 400.000 pages takes normally 12secs. after the last upgrade of the kernel (wheezy or sid) it takes 40 secs. It is not the disk space because when I move that to a different device, the same thing happens. This might be a feature, but it is so noticable :-( Thanks! Is that reproducible without the nvidia driver? Yes, Julien, behavior is the same slow hibernate w/o the nvidia driver. Hugo -- Being unified, you have no liking; being transmuted, you have no fixity. http://gpc-qt.sourceforge.net/ https://linuxcounter.net/machine/419602.html http://debian-oaxaca.blogspot.com/
Bug#683768: linux-image-3.2.0-3-amd64: hibernate 3x slower after upgrade to linux-image-3.2.0-3-amd64
On Fri, Aug 03, 2012 at 02:09:38PM -0500, hugo vanwoerkom wrote: On Fri, Aug 3, 2012 at 12:36 PM, Julien Cristau jcris...@debian.org wrote: On Fri, Aug 3, 2012 at 12:06:13 -0500, hugo vanwoerkom wrote: Package: src:linux Version: 3.2.23-1 Severity: normal Hibernate of about 400.000 pages takes normally 12secs. after the last upgrade of the kernel (wheezy or sid) it takes 40 secs. You don't say what the previous version was. You may be able to find this information in /var/log/dpkg.log if you don't remember. It is not the disk space because when I move that to a different device, the same thing happens. This might be a feature, but it is so noticable :-( Thanks! Is that reproducible without the nvidia driver? Yes, Julien, behavior is the same slow hibernate w/o the nvidia driver. There are some fixes for bad hibernation performance in Linux 3.2.24. You can try these out by applying http://git.kernel.org/?p=linux/kernel/git/bwh/linux-3.2.y.git;a=patch;h=6ed6791a1697afcb1615b4252d0c304a743b5f4d following the instructions at http://kernel-handbook.alioth.debian.org/ch-common-tasks.html#s4.2.2. Ben. -- Ben Hutchings We get into the habit of living before acquiring the habit of thinking. - Albert Camus -- To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20120803202203.gi1...@decadent.org.uk
Bug#683768: linux-image-3.2.0-3-amd64: hibernate 3x slower after upgrade to linux-image-3.2.0-3-amd64
On Fri, Aug 3, 2012 at 3:22 PM, Ben Hutchings b...@decadent.org.uk wrote: On Fri, Aug 03, 2012 at 02:09:38PM -0500, hugo vanwoerkom wrote: On Fri, Aug 3, 2012 at 12:36 PM, Julien Cristau jcris...@debian.org wrote: On Fri, Aug 3, 2012 at 12:06:13 -0500, hugo vanwoerkom wrote: Package: src:linux Version: 3.2.23-1 Severity: normal Hibernate of about 400.000 pages takes normally 12secs. after the last upgrade of the kernel (wheezy or sid) it takes 40 secs. You don't say what the previous version was. You may be able to find this information in /var/log/dpkg.log if you don't remember. Sorry, the previous version was 3.2.19-1. Hugo It is not the disk space because when I move that to a different device, the same thing happens. This might be a feature, but it is so noticable :-( Thanks! Is that reproducible without the nvidia driver? Yes, Julien, behavior is the same slow hibernate w/o the nvidia driver. There are some fixes for bad hibernation performance in Linux 3.2.24. You can try these out by applying http://git.kernel.org/?p=linux/kernel/git/bwh/linux-3.2.y.git;a=patch;h=6ed6791a1697afcb1615b4252d0c304a743b5f4d following the instructions at http://kernel-handbook.alioth.debian.org/ch-common-tasks.html#s4.2.2. Ben. -- Ben Hutchings We get into the habit of living before acquiring the habit of thinking. - Albert Camus
Bug#683768: linux-image-3.2.0-3-amd64: hibernate 3x slower after upgrade to linux-image-3.2.0-3-amd64
On Fri, 2012-08-03 at 16:24 -0500, hugo vanwoerkom wrote: On Fri, Aug 3, 2012 at 3:22 PM, Ben Hutchings b...@decadent.org.uk wrote: On Fri, Aug 03, 2012 at 02:09:38PM -0500, hugo vanwoerkom wrote: On Fri, Aug 3, 2012 at 12:36 PM, Julien Cristau jcris...@debian.org wrote: On Fri, Aug 3, 2012 at 12:06:13 -0500, hugo vanwoerkom wrote: Package: src:linux Version: 3.2.23-1 Severity: normal Hibernate of about 400.000 pages takes normally 12secs. after the last upgrade of the kernel (wheezy or sid) it takes 40 secs. You don't say what the previous version was. You may be able to find this information in /var/log/dpkg.log if you don't remember. Sorry, the previous version was 3.2.19-1. I can't see any change since 3.2.19-1 that would explain this. But please do test the patch: There are some fixes for bad hibernation performance in Linux 3.2.24. You can try these out by applying http://git.kernel.org/?p=linux/kernel/git/bwh/linux-3.2.y.git;a=patch;h=6ed6791a1697afcb1615b4252d0c304a743b5f4d following the instructions at http://kernel-handbook.alioth.debian.org/ch-common-tasks.html#s4.2.2. -- Ben Hutchings Experience is directly proportional to the value of equipment destroyed. - Carolyn Scheppner signature.asc Description: This is a digitally signed message part
Bug#683768: linux-image-3.2.0-3-amd64: hibernate 3x slower after upgrade to linux-image-3.2.0-3-amd64
On Fri, Aug 3, 2012 at 6:59 PM, Ben Hutchings b...@decadent.org.uk wrote: On Fri, 2012-08-03 at 16:24 -0500, hugo vanwoerkom wrote: On Fri, Aug 3, 2012 at 3:22 PM, Ben Hutchings b...@decadent.org.uk wrote: On Fri, Aug 03, 2012 at 02:09:38PM -0500, hugo vanwoerkom wrote: On Fri, Aug 3, 2012 at 12:36 PM, Julien Cristau jcris...@debian.org wrote: On Fri, Aug 3, 2012 at 12:06:13 -0500, hugo vanwoerkom wrote: Package: src:linux Version: 3.2.23-1 Severity: normal Hibernate of about 400.000 pages takes normally 12secs. after the last upgrade of the kernel (wheezy or sid) it takes 40 secs. You don't say what the previous version was. You may be able to find this information in /var/log/dpkg.log if you don't remember. Sorry, the previous version was 3.2.19-1. I can't see any change since 3.2.19-1 that would explain this. But please do test the patch: Running the compile right now. Will let you know. Thanks Ben! Hugo There are some fixes for bad hibernation performance in Linux 3.2.24. You can try these out by applying http://git.kernel.org/?p=linux/kernel/git/bwh/linux-3.2.y.git;a=patch;h=6ed6791a1697afcb1615b4252d0c304a743b5f4d following the instructions at http://kernel-handbook.alioth.debian.org/ch-common-tasks.html#s4.2.2. -- Ben Hutchings Experience is directly proportional to the value of equipment destroyed. - Carolyn Scheppner