Possible deadlock with -writepaged version of flush_dirty_buffers()and 2.4.0

2001-01-10 Thread Marcelo Tosatti
Hi Chris, It seems there is a possible deadlock condition with your patch which changes flush_dirty_buffers() to use -writepage (something which we _definately_ want for 2.5). Take a look: mark_buffer_dirty-balance_dirty-wakeup_bdflush-flush_dirty_buffers-

Re: 2.4.0 vm BUG (ksymoopsed)

2001-01-10 Thread Marcelo Tosatti
On Thu, 11 Jan 2001, Ulrich Schwarz wrote: 2.4.0 (final i586) patched with reiserfs 3.6.25 produced the following BUG: the console report (ksymoopsed): kernel BUG at vmscan.c:452! invalid operand: Does reiserfs patch changes vmscan.c ? If so, whats in line 452 of mm/vmscan.c of

set_page_dirty/page_launder deadlock

2001-01-13 Thread Marcelo Tosatti
Hi, While taking a look at page_launder()... /* And re-start the thing.. */ spin_lock(pagemap_lru_lock); -- if (result != 1) continue; /* writepage refused to do anything */

Re: mmap()/VM problem in 2.4.0

2001-01-14 Thread Marcelo Tosatti
On Fri, 12 Jan 2001, Vlad Bolkhovitine wrote: After upgrade from 2.4.0-test7 to 2.4.0 while running tiotest v0.3.1 I found two following problems. There have been quite a lot of things changed from 2.4.0-test7 to 2.4.0, so I'm not sure what caused the slowdown. Anyway, important VM changes

Re: 2.4.0-ac9 works, but slower and swappier

2001-01-14 Thread Marcelo Tosatti
On Sun, 14 Jan 2001, Mark Orr wrote: I've been running 2.4.0-ac9 for a day and a half now. I have pretty low-end hardware (Pentium 1/ 100MHz, 16Mb RAM, 17Mb swap) and it really seems to bog down with anything heavy in memory.Netscape seems to really drag, and any Java applets I

Re: FS corruption on 2.4.0-ac8

2001-01-15 Thread Marcelo Tosatti
On Mon, 15 Jan 2001, Jure Pecar wrote: Hi all, I was running 2.4.0test10pre5 happily for months and wanted to see how things stand in the 'latest stuff'. Here's what i found: I compiled 2.4.0-ac8 with nearly the same .config as test10pre5 (with latest gcc on rh7). Then i booted it and

Re: Oops with 4GB memory setting in 2.4.0 stable

2001-01-15 Thread Marcelo Tosatti
On Tue, 16 Jan 2001, Rainer Mager wrote: Attached is my oops.txt and the result sent through ksymoops. The results don't look particularly useful to me so perhaps I'm doing something wrong. PLEASE tell me if I should parse this differently. Likewise, if there is anything else I can

RE: Oops with 4GB memory setting in 2.4.0 stable

2001-01-15 Thread Marcelo Tosatti
On Tue, 16 Jan 2001, Rainer Mager wrote: I knew that, I was just testing you all. ;-) EIP; f889e044 END_OF_CODE+385bfe34/ = Trace; f889d966 END_OF_CODE+385bf756/ Trace; c0140c10 vfs_readdir+90/ec Trace; c0140e7c filldir+0/d8 Trace; c0140f9e sys_getdents+4a/98 Trace; c0140e7c

RE: Oops with 4GB memory setting in 2.4.0 stable

2001-01-15 Thread Marcelo Tosatti
On Tue, 16 Jan 2001, Rainer Mager wrote: Ok, now were making progress. I did as you said and have attached (really!) the new parsed output. Now we have some useful information (I hope). I still got lots of warnings on symbols (which I have edited out of the parsed file for the sake of

2.4.1pre8 slowdown on dbench tests

2001-01-18 Thread Marcelo Tosatti
Hi, On my dbench runs I've noted a slowdown between pre4 and pre8 with 48 threads. (128MB, 2 CPU's machine) pre4: Throughput 7.05841 MB/sec (NB=8.82301 MB/sec 70.5841 MBit/sec) 70.94user 232.54system 15:17.39elapsed 33%CPU (0avgtext+0avgdata 0maxresident)k 0inputs+0outputs

Re: mmap()/VM problem in 2.4.0

2001-01-18 Thread Marcelo Tosatti
On Thu, 18 Jan 2001, Rik van Riel wrote: On Fri, 12 Jan 2001, Vlad Bolkhovitine wrote: You can see, mmap() read performance dropped significantly as well as read() one raised. Plus, "interactivity" of 2.4.0 system was much worse during mmap'ed test, than using read() (everything was

Re: 2.4.1pre8 slowdown on dbench tests

2001-01-18 Thread Marcelo Tosatti
On Thu, 18 Jan 2001, Steven Cole wrote: On Thu, 18 Jan 2001, Marcelo Tosatti wrote: On my dbench runs I've noted a slowdown between pre4 and pre8 with 48 threads. (128MB, 2 CPU's machine) I ran dbench 48 four times in succession for 2.4.0 and 2.4.1-pre8. The change in performance

Re: 2.4.1pre8 slowdown on dbench tests

2001-01-18 Thread Marcelo Tosatti
On Fri, 19 Jan 2001, Andrea Arcangeli wrote: Marcelo can you give a try with `high_queued_sectors = total_ram / 3' and low_queued_sectors = high_queued_sectors / 2 and drop the big ram machine check? Andrea, With the changes you suggested I got almost the same results with pre8. - To

Re: pre5 VM feedback..

2001-01-18 Thread Marcelo Tosatti
On 15 Jan 2001, Linus Torvalds wrote: In article [EMAIL PROTECTED], Jeff Garzik [EMAIL PROTECTED] wrote: $!@#@! pre6 is already out :) Yes, and for heavens sake don't use it, because the reiserfs merge got some dirty inode logic wrong. pre7 fixes just that one line and should be ok

Re: pre5 VM feedback..

2001-01-18 Thread Marcelo Tosatti
On Fri, 19 Jan 2001, Marcelo Tosatti wrote: The swapin readahead code tries to allocate (1 page_cluster) pages at each swapin. This means there's a big chance of having (1 page_cluster) "self-swap-out"'s at each swapin if we're under low memory. Nasty. Actually its

[RFC] generic IO write clustering

2001-01-19 Thread Marcelo Tosatti
Hi, I'm starting to implement a generic write clustering scheme and I would like to receive comments and suggestions. The write clustering issue has already been discussed (mainly at Miami) and the agreement, AFAIK, was to implement the write clustering at the per-address-space writepage()

Re: [RFC] generic IO write clustering

2001-01-19 Thread Marcelo Tosatti
On Sat, 20 Jan 2001, Rik van Riel wrote: Is there ever a reason NOT to do the best possible IO clustering at write time ? Remember that disk writes do not cost memory and have no influence on the resident set ... completely unlike read clustering, which does need to be limited. You dont

Re: [RFC] generic IO write clustering

2001-01-20 Thread Marcelo Tosatti
On Sat, 20 Jan 2001, Christoph Hellwig wrote: snip I think there is a big disadvantage of this appropeach: To find out which pages are clusterable, we need do do bmap/get_block, that means we have to go through the block-allocation functions, which is rather expensive, and then we have to

Re: [RFC] generic IO write clustering

2001-01-20 Thread Marcelo Tosatti
On Sat, 20 Jan 2001, Christoph Hellwig wrote: On Sat, Jan 20, 2001 at 01:24:40PM -0200, Marcelo Tosatti wrote: In case the metadata was not already cached before -cluster() (in this case there is no disk IO at all), -cluster() will cache it avoiding further disk accesses by writepage

Re: [RFC] generic IO write clustering

2001-01-20 Thread Marcelo Tosatti
On Sat, 20 Jan 2001, Christoph Hellwig wrote: On Sat, Jan 20, 2001 at 02:00:24PM -0200, Marcelo Tosatti wrote: True. But you have to go through ext2_get_branch (under the big kernel lock) - if we can do only one logical-physical block translations, why doing it multiple times

Re: mmap()/VM problem in 2.4.0

2001-01-21 Thread Marcelo Tosatti
On Fri, 19 Jan 2001, Rik van Riel wrote: On Thu, 18 Jan 2001, Marcelo Tosatti wrote: On Thu, 18 Jan 2001, Rik van Riel wrote: On Fri, 12 Jan 2001, Vlad Bolkhovitine wrote: You can see, mmap() read performance dropped significantly as well as read() one raised. Plus

Re: 2.4.1pre8 slowdown on dbench tests

2001-01-22 Thread Marcelo Tosatti
Jens, Steven is a seeing a slowdown in his results, too. On Mon, 22 Jan 2001, Steven Cole wrote: On Thursday 18 January 2001 14:49, Marcelo Tosatti wrote: Steven, The issue is the difference between pre4 and pre8. Could you please try pre4 and report results ? Thanks Ok

Re: 2.4.1-test10

2001-01-22 Thread Marcelo Tosatti
Any technical reason why the background page aging fix was not applied? On Mon, 22 Jan 2001, Linus Torvalds wrote: The ChangeLog may not be 100% complete. The physically big things are the PPC and ACPI updates, even if most people won't notice. - To unsubscribe from this list: send the line

Re: __alloc_pages: 3-order allocation failed. for 2.4.1-pre9

2001-01-23 Thread Marcelo Tosatti
On 23 Jan 2001, Yann Dupont wrote: I remember sawing that those errors were due to improperly written drivers . Is the buslogic driver or tape driver are to blame here ?? Or maybe this is a vm balancing issue ? Could you please send the output of "Alt+SysRq+m" (kernel must be compiled with

Re: 2.4.1-test10

2001-01-23 Thread Marcelo Tosatti
On Tue, 23 Jan 2001, Andre Hedrick wrote: Just my nickel on the issue. Andre, This patch I'm talking about is for a different issue from what was discussed in the IO clustering thread. - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to

Re: inode-i_dirty_buffers redundant ?

2001-01-24 Thread Marcelo Tosatti
On Wed, 24 Jan 2001, V Ganesh wrote: now that we have inode-i_mapping-dirty_pages, what do we need inode-i_dirty_buffers for ? I understand the latter was added for the O_SYNC changes before dirty_pages came into the picture. but now both seem to be doing more or less the same thing.

Re: inode-i_dirty_buffers redundant ?

2001-01-25 Thread Marcelo Tosatti
On Thu, 25 Jan 2001, Stephen C. Tweedie wrote: Hi, On Thu, Jan 25, 2001 at 04:17:30PM +0530, V Ganesh wrote: so i_dirty_buffers contains buffer_heads of pages coming from write() as well as metadata buffers from mark_buffer_dirty_inode(). a dirty MAP_SHARED page which has been

Re: inode-i_dirty_buffers redundant ?

2001-01-25 Thread Marcelo Tosatti
On Thu, 25 Jan 2001, Daniel Phillips wrote: "Stephen C. Tweedie" wrote: We also maintain the per-page buffer lists as caches of the virtual-to-physical mapping to avoid redundant bmap()ping. Could you clarify that one, please? Daniel, With "physical mapping" Stephen means on-disk

Re: VM breakdown, 2.4.0 family

2001-01-27 Thread Marcelo Tosatti
On Sat, 27 Jan 2001, David Ford wrote: Since the testN series and up through ac12, I experience total loss of control when memory is nearly exhausted. I start with 256M and eat it up with programs until there is only about 7 megs left, no swap. From that point all user processes stall

Re: VM breakdown, 2.4.0 family

2001-01-27 Thread Marcelo Tosatti
On Sat, 27 Jan 2001, David Ford wrote: I have Marcelo's patch. It isn't applicable because I am purposely not enabling any swap. The problem is the system gets down to about 7 megs of buffers free and within three seconds has become functionally dead. Zero response on any user

Re: 2.4.1-pre10 deadlock (Re: ps hang in 241-pre10)

2001-01-27 Thread Marcelo Tosatti
On Sun, 28 Jan 2001, Jens Axboe wrote: On Sat, Jan 27 2001, Linus Torvalds wrote: What was the trace of this? Just curious, the below case outlined by Linus should be pretty generic, but I'd still like to know what can lead to this condition. It was posted on linux-kernel - I

Re: ps hang in 241-pre10

2001-01-27 Thread Marcelo Tosatti
(ugh, sorry about last mail) On 27 Jan 2001, Linus Torvalds wrote: In article [EMAIL PROTECTED], David Ford [EMAIL PROTECTED] wrote: Unfortunately klogd reads /procerg. So the following is a painstakingly slow hand translation, I'll only print the D state entries unless someone asks

Re: ps hang in 241-pre10

2001-01-27 Thread Marcelo Tosatti
On Sat, 27 Jan 2001, Linus Torvalds wrote: On Sun, 28 Jan 2001, Marcelo Tosatti wrote: This is the smoking gun here, I bet, but I'd like to make sure I see the whole thing. I don't see _why_ we'd have deadlocked on __wait_on_page(), but I think this is the thread that hangs

Re: ps hang in 241-pre10

2001-01-27 Thread Marcelo Tosatti
On Sun, 28 Jan 2001, Marcelo Tosatti wrote: On Sat, 27 Jan 2001, Linus Torvalds wrote: On Sun, 28 Jan 2001, Marcelo Tosatti wrote: This is the smoking gun here, I bet, but I'd like to make sure I see the whole thing. I don't see _why_ we'd have deadlocked

Re: [PATCH] 2.4.1 find_page_nolock fixes

2001-01-30 Thread Marcelo Tosatti
On Tue, 30 Jan 2001, Rik van Riel wrote: Hi Linus, the patch below contains 3 small changes to mm/filemap.c: 1. replace the aging in __find_page_nolock() with setting PageReferenced(), otherwise a large number of small reads from (or writes to) a page can drive up the page

[PATCH] vma limited swapin readahead

2001-01-30 Thread Marcelo Tosatti
Hi, The current swapin readahead code reads a number of pages (1 page_cluster) which are physically contiguous on disk with reference to the page which needs to be faulted in. However, the pages which are contiguous on swap are not necessarily contiguous in the virtual memory area where the

Re: [PATCH] vma limited swapin readahead

2001-01-31 Thread Marcelo Tosatti
On Wed, 31 Jan 2001, Stephen C. Tweedie wrote: Hi, On Wed, Jan 31, 2001 at 01:05:02AM -0200, Marcelo Tosatti wrote: However, the pages which are contiguous on swap are not necessarily contiguous in the virtual memory area where the fault happened. That means the swapin readahead

Re: kernel BUG at inode.c:889!

2001-01-31 Thread Marcelo Tosatti
On Wed, 31 Jan 2001, Timo Jantunen wrote: Heip! While I was looking unused partitions to be used for ReiserFS testing Haven't you forgot to inform which kernel version are you using? - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to

Re: [PATCH] allocation looping + kswapd CPU cycles

2001-05-10 Thread Marcelo Tosatti
On Thu, 10 May 2001, Mark Hemment wrote: On Wed, 9 May 2001, Marcelo Tosatti wrote: On Wed, 9 May 2001, Mark Hemment wrote: Could introduce another allocation flag (__GFP_FAIL?) which is or'ed with a __GFP_WAIT to limit the looping? __GFP_FAIL is in the -ac tree already

Re: [PATCH] allocation looping + kswapd CPU cycles

2001-05-10 Thread Marcelo Tosatti
On Thu, 10 May 2001, Stephen C. Tweedie wrote: Hi, On Thu, May 10, 2001 at 01:43:46PM -0300, Marcelo Tosatti wrote: No. __GFP_FAIL can to try to reclaim pages from inactive clean. We just want to avoid __GFP_FAIL allocations from going to try_to_free_pages(). Why? __GFP_FAIL

Re: [PATCH] allocation looping + kswapd CPU cycles

2001-05-10 Thread Marcelo Tosatti
On Thu, 10 May 2001, Stephen C. Tweedie wrote: Hi, On Thu, May 10, 2001 at 03:22:57PM -0300, Marcelo Tosatti wrote: Initially I thought about __GFP_FAIL to be used by writeout routines which want to cluster pages until they can allocate memory without causing any pressure

[PATCH] VM fixes against 2.4.4-ac6

2001-05-11 Thread Marcelo Tosatti
Hi, The following patch addresses two issues: - Buffer cache pages in the inactive lists are not getting their age increased if they get touched by getblk (which will set the referenced bit on the page). page_launder() simply cleans the referenced bit on such pages and moves them to the

Re: [PATCH] VM fixes against 2.4.4-ac6

2001-05-11 Thread Marcelo Tosatti
On Fri, 11 May 2001, Marcelo Tosatti wrote: Hi, The following patch addresses two issues: - Buffer cache pages in the inactive lists are not getting their age increased if they get touched by getblk (which will set the referenced bit on the page). page_launder() simply cleans

Re: [PATCH] writepage method changes

2001-05-10 Thread Marcelo Tosatti
On Wed, 9 May 2001, Marcelo Tosatti wrote: Locked for the not wrote out case (I will fix my patch now, thanks) I just found out that there are filesystems (eg reiserfs) which write out data even if an error ocurred, which means the unlocking must be done by the filesystems, always

[PATCH] writepage method changes

2001-05-09 Thread Marcelo Tosatti
Well, Here is the updated version of the patch to add the priority argument to writepage(). All implementations have been fixed. No referenced bit changes as I still think its not worth passing this information down to writepage(). Note: I've removed ramfs_writepage(). If there is no

Re: [PATCH] writepage method changes

2001-05-10 Thread Marcelo Tosatti
On Thu, 10 May 2001, Chris Mason wrote: On Wednesday, May 09, 2001 10:51:17 PM -0300 Marcelo Tosatti [EMAIL PROTECTED] wrote: On Wed, 9 May 2001, Marcelo Tosatti wrote: Locked for the not wrote out case (I will fix my patch now, thanks) I just found out

Re: page_launder() bug

2001-05-09 Thread Marcelo Tosatti
On Wed, 9 May 2001, David S. Miller wrote: Marcelo Tosatti writes: You want writepage() to check/clean the referenced bit and move the page to the active list itself ? Well, that's the other part of what my patch was doing. Let me state it a different way, how is the new

Re: [PATCH] writepage method changes

2001-05-10 Thread Marcelo Tosatti
On Thu, 10 May 2001, Andrew Morton wrote: Marcelo Tosatti wrote: Well, Here is the updated version of the patch to add the priority argument to writepage(). It appears that a -EIO return from block_write_full_page() will result in an unlock of an unlocked page in page_launder

Re: page_launder() bug

2001-05-14 Thread Marcelo Tosatti
On Sun, 13 May 2001, Linus Torvalds wrote: On Sun, 13 May 2001, Rik van Riel wrote: This means that the swapin path (and the same path for other pagecache pages) doesn't take the page lock and the page lock doesn't protect us from other people using the page while we have it

Re: [PATCH] vmscan.c fixes

2001-05-14 Thread Marcelo Tosatti
On Mon, 14 May 2001, Rik van Riel wrote: Hi Linus, the following patch does: snip pg_data_t *pgdat = pgdat_list; int sum = 0; int freeable = nr_free_pages() + nr_inactive_clean_pages(); + /* XXX: dynamic free target is complicated and may be wrong... */

Re: [PATCH] v2.4.4-ac9 highmem deadlock

2001-05-14 Thread Marcelo Tosatti
On Mon, 14 May 2001, Ben LaHaise wrote: Hey folks, Hi. The patch below consists of 3 seperate fixes for helping remove the deadlocks present in current kernels with respect to highmem systems. Each fix is to a seperate file, so please accept/reject as such. snip The third patch (to

[PATCH] remove page_launder() from bdflush

2001-05-15 Thread Marcelo Tosatti
Hi Linus, There is no reason why bdflush should call page_launder(). Its pretty obvious that bdflush's job is to only write out _buffers_. Under my tests this patch makes things faster. Guess why? Because bdflush is writing out buffers when it should instead blocking inside

Re: mm/memory.c: Missing pte_mkyoung() on mk_pte() calls?

2001-05-17 Thread Marcelo Tosatti
Two seconds after I sent the message Benjamin told me on IRC that PAGE_ACCESSED is included in the default page protections... duh. On Thu, 17 May 2001, Marcelo Tosatti wrote: Linus, I was looking at mm/memory.c (2.4), and I've noticed that we don't call pte_mkyoung() on newly created

mm/memory.c: Missing pte_mkyoung() on mk_pte() calls?

2001-05-17 Thread Marcelo Tosatti
Linus, I was looking at mm/memory.c (2.4), and I've noticed that we don't call pte_mkyoung() on newly created pte's for most of the fault paths. break_cow(), for example: establish_pte(vma, address, page_table, pte_mkwrite(pte_mkdirty(mk_pte(new_page, v ma-vm_page_prot; Is there any

Re: page_launder() bug

2001-05-09 Thread Marcelo Tosatti
On Wed, 9 May 2001, David S. Miller wrote: Marcelo Tosatti writes: Let me state it a different way, how is the new writepage() framework going to do things like ignore the referenced bit during page_launder for dead swap pages? Its not able to ignore the referenced bit

Re: [RFC][PATCH] Re: Linux 2.4.4-ac10

2001-05-20 Thread Marcelo Tosatti
On Sun, 20 May 2001, Mike Galbraith wrote: Also in all recent kernels, if the machine is swapping, swap cache grows without limits and is hard to recycle, but then again that is a known problem. This one bugs me. I do not see that and can't understand why. To throw away dirty and

Re: [RFC][PATCH] Re: Linux 2.4.4-ac10

2001-05-20 Thread Marcelo Tosatti
On Sat, 19 May 2001, Mike Galbraith wrote: @@ -1054,7 +1033,7 @@ if (!zone-size) continue; - while (zone-free_pages zone-pages_low) { + while (zone-free_pages

write drop behind effect on active scanning

2001-05-23 Thread Marcelo Tosatti
Hi, I just noticed a bad effect of write drop behind yesterday during some tests. The problem is that we deactivate written pages, thus making the inactive list become pretty big (full of unfreeable pages) under write intensive IO workloads. So what happens is that we don't do _any_ aging

Re: write drop behind effect on active scanning

2001-05-23 Thread Marcelo Tosatti
On Wed, 23 May 2001, Daniel Phillips wrote: On Wednesday 23 May 2001 09:33, Marcelo Tosatti wrote: Hi, I just noticed a bad effect of write drop behind yesterday during some tests. The problem is that we deactivate written pages, thus making the inactive list become pretty big

Re: [patch] 4GB I/O, 2nd edition

2001-05-28 Thread Marcelo Tosatti
On Mon, 28 May 2001, Jens Axboe wrote: Hi, One minor bug found that would possibly oops if the SCSI pool ran out of memory for the sg table and had to revert to a single segment request. This should never happen, as the pool is sized after number of devices and queue depth -- but it

Re: Linux 2.4.5-ac2

2001-05-28 Thread Marcelo Tosatti
On Tue, 29 May 2001, André Dahlqvist wrote: André Dahlqvist [EMAIL PROTECTED] wrote: I agree. Kernels after 2.4.4 uses a *lot* more swap for me, which I guess might be part of the reason for the slowdown. Following up on myself, here are some numbers: Freshly booted 2.4.4 with X

Re: Plain 2.4.5 VM...

2001-05-29 Thread Marcelo Tosatti
On Tue, 29 May 2001, Jakob Østergaard wrote: It's not a bug. It's a feature. It only breaks systems that are run with too little swap, and the only difference from 2.2 till now is, that the definition of too little changed. Its just a balancing change, actually. You can tune the code

Re: Plain 2.4.5 VM

2001-05-30 Thread Marcelo Tosatti
On Wed, 30 May 2001, Jonathan Morton wrote: The page aging logic does seems fragile as heck. You never know how many folks are aging pages or at what rate. If aging happens too fast, it defeats the garbage identification logic and you rape your cache. If aging happens too slowly..

Re: Zerocopy NBD

2001-05-30 Thread Marcelo Tosatti
On Wed, 30 May 2001, Steve Whitehouse wrote: Hi, Attached is a patch I came up with recently to do add zerocopy support to NBD for writes. I'm not intending that this should go into the kernel before at least 2.5, I'm just sending it here in case it is useful to anyone. I wrote it is

Re: Zerocopy NBD

2001-05-30 Thread Marcelo Tosatti
On Wed, 30 May 2001, Steve Whitehouse wrote: Hi, On Wed, 30 May 2001, Steve Whitehouse wrote: [info about NBD patch deleted] Cool. Are you seeing performance improvements with the patch ? Yes, but my testing is not in anyway complete yet. The only network device

Re: Plain 2.4.5 VM

2001-05-30 Thread Marcelo Tosatti
On Wed, 30 May 2001, Rik van Riel wrote: On Wed, 30 May 2001, Marcelo Tosatti wrote: The problem is that we allow _every_ task to age pages on the system at the same time --- this is one of the things which is fucking up. This should not have any effect on the ratio of cache

Re: Plain 2.4.5 VM

2001-05-30 Thread Marcelo Tosatti
On Wed, 30 May 2001, Mike Galbraith wrote: On Wed, 30 May 2001, Rik van Riel wrote: On Wed, 30 May 2001, Marcelo Tosatti wrote: The problem is that we allow _every_ task to age pages on the system at the same time --- this is one of the things which is fucking up. This should

Re: [PATCH] reclaim dirty dead swapcache pages

2001-05-30 Thread Marcelo Tosatti
On Thu, 31 May 2001, J . A . Magallon wrote: On 05.30 Marcelo Tosatti wrote: Its at http://bazar.conectiva.com.br/~marcelo/patches/v2.4/2.4.5ac4/reapswap.patch Please test. Which kind of test, something like the gcc think I posted recently ? I don't remember

Re: Linux-2.4.5

2001-05-27 Thread Marcelo Tosatti
On Sat, 26 May 2001, Marcelo Tosatti wrote: You're trying to fix the symptoms, by attacking the final end. And what I've been trying to say is that this problem likely has a higher-level _cause_, and I want that _cause_ fixed. Not the symptoms. You are not going to fix the problem

Comment on patch to remove nr_async_pages limit

2001-06-04 Thread Marcelo Tosatti
Zlatko, I've read your patch to remove nr_async_pages limit while reading an archive on the web. (I have to figure out why lkml is not being delivered correctly to me...) Quoting your message: That artificial limit hurts both swap out and swap in path as it introduces synchronization

Re: 2.2.18-3 seems to break pcmcia-cs-3.1.20

2000-09-05 Thread Marcelo Tosatti
On Tue, 5 Sep 2000 [EMAIL PROTECTED] wrote: # make all make[1]: Entering directory /usr/src/pcmcia-cs-3.1.20/modules' cc -MD -O2 -Wall -Wstrict-prototypes -pipe -I../include -I/usr/src/linux/include -D__KERNEL__ -DMODULE -c cs.c In file included from

Re: 2.2.18pre2aa2 and patches for 2.2.18pre3

2000-09-07 Thread Marcelo Tosatti
On Thu, 7 Sep 2000, Matthew Hawkins wrote: I'd like to advocate the inclusion of the majority of these patches of Andrea's. I've been patching most of them in for a while now simply because I've found my SMP system much more stable and useable. Andrea VM patches will be included in

Re: Error in fs/nls/Config.in in 2.2.18-pre3

2000-09-07 Thread Marcelo Tosatti
On Thu, 7 Sep 2000, Urban Widmark wrote: On Thu, 7 Sep 2000, G. Hugh Song wrote: if [ "$CONFIG_JOLIET" = "y" -o "$CONFIG_FAT_FS" != "n" \ -o "$CONFIG_NTFS_FS" != "n" -o "$CONFIG_NCPFS_NLS" = "y" \ -o "$CONFIG_SMB_FS" != n ]; then n vs "n" is my error. However

Re: Error in fs/nls/Config.in in 2.2.18-pre3

2000-09-07 Thread Marcelo Tosatti
On Fri, 8 Sep 2000, Urban Widmark wrote: On Thu, 7 Sep 2000, Marcelo Tosatti wrote: oldconfig always ask about CONFIG_SMB_NLS_REMOTE in case it was set CONFIG_SMB_NLS_REMOTE="" in the previous config. This is expected? It's certainly annoying, especially for peopl

Re: [PATCH] af_netrom.c: do resource release on failure

2000-09-08 Thread Marcelo Tosatti
On Fri, 8 Sep 2000, Torben Mathiasen wrote: On Fri, Sep 08 2000, Arnaldo Carvalho de Melo wrote: Hi, Please take a look and consider applying. Some of it are small cleanups, if they're deemed unnecessary, lemme now and I'm back it off. I think that there are some more unchecked

Re: Oops on boot with both 2.2.17 and 2.4.0t8p6

2000-09-09 Thread Marcelo Tosatti
On Sat, 9 Sep 2000, Rasmus Andersen wrote: snip Code: 0f b6 0c 03 89 4c 24 14 51 68 8e e5 17 c0 e8 de a4 00 00 83 EIP; c0107f27 show_registers+237/268 = Trace; c300 END_OF_CODE+2e30398/ Trace; c0107f85 die+2d/38 Thats not the first oops yet, and as Keith told you, its

Re: An elevator algorithm (patch)

2000-09-17 Thread Marcelo Tosatti
On Sun, 17 Sep 2000, Andrea Arcangeli wrote: snip If nobody does that before me I will try this "remeber last position of the head" idea in my blkdev tree (there are many other pending elevator fixes in it) as soon as I finished with 2.2.18pre9aa1 LFS nfsv3 and as soon as I finish the fix

Re: Oops while using bonded ethernet

2000-09-22 Thread Marcelo Tosatti
On Fri, 22 Sep 2000, Michael R. Jinks wrote: I'm trying to bond all four interfaces of a D-Link DFE-570TX ethernet card. Not sure who maintains the bonding module, so writing directly to the main list. Tips on better people to bother are welcome. Note on my kernel version: I'm using

Re: Given an image, how can show its config?

2000-09-23 Thread Marcelo Tosatti
On Sat, 23 Sep 2000, David Ford wrote: Keith Owens wrote: That would take my 2.4.0 bzImage to 893864, it does not leave much room out of a 1.4Mb floppy for LILO files. We could have multiple make targets, with and without appended config/map but that just complicates the build

Re: [patch] vmfixes-2.4.0-test9-B2

2000-09-24 Thread Marcelo Tosatti
On Sun, 24 Sep 2000, Linus Torvalds wrote: On Sun, 24 Sep 2000, Andrea Arcangeli wrote: On Sun, Sep 24, 2000 at 10:26:11PM +0200, Ingo Molnar wrote: where will it deadlock? ext2_new_block (or whatever that runs getblk with the superlock lock

Re: [patch] vmfixes-2.4.0-test9-B2

2000-09-24 Thread Marcelo Tosatti
On Mon, 25 Sep 2000, Andrea Arcangeli wrote: snip kmem_cache_reap shrinks the slabs at _very_ low frequency. It's worthless to keep lots of dentries and icache into the slab internal queues until kmem_cache_reap kicks in again, if we free them such memory immediatly instead we'll run

Re: more testing on 2.4.0-t9p[456] VM deadlocks

2000-09-25 Thread Marcelo Tosatti
On Mon, 25 Sep 2000, Martin Diehl wrote: On Mon, 25 Sep 2000, Martin Diehl wrote: PS: vmfixes-2.4.0-test9-B2 not yet tested - will do later. Hi - done now: using 2.4.0-t9p6 + vmfixes-2.4.0-test9-B2 I ended up with the box deadlocked again! Was "make bzImage" on UP booted with

Re: the new VM

2000-09-25 Thread Marcelo Tosatti
On Mon, 25 Sep 2000, Andrea Arcangeli wrote: snip I talked with Alexey about this and it seems the best way is to have a per-socket reservation of clean cache in function of the receive window. So we don't need an huge atomic pool but we can have a special lru with an irq spinlock that is

Re: more testing on 2.4.0-t9p[456] VM deadlocks

2000-09-26 Thread Marcelo Tosatti
On Tue, 26 Sep 2000, Ingo Molnar wrote: On 26 Sep 2000, Juan J. Quintela wrote: Ingo, I am very wrong, or vmfixes-B2_deadlock is not included in test9-pre7. well, the __GFP_IO part is included (in a different way). The slab.c part is not included. Actually the __GFP_IO

Re: what causes Machine Check exception?

2000-09-26 Thread Marcelo Tosatti
Alan, I think adding a document about MCE in the kernel would be very useful. Or at least a pointer to Intel's documentation about it. On 26 Sep 2000, H. Peter Anvin wrote: Followup to: [EMAIL PROTECTED] By author:"Martin Bene" [EMAIL PROTECTED] In newsgroup: linux.dev.kernel

Re: more testing on 2.4.0-t9p[456] VM deadlocks

2000-09-27 Thread Marcelo Tosatti
On Wed, 27 Sep 2000, Ingo Molnar wrote: On Tue, 26 Sep 2000, Marcelo Tosatti wrote: well, the __GFP_IO part is included (in a different way). The slab.c part is not included. Actually the __GFP_IO check is now only inside ext2. no, it isnt. It's in the VFS. In fact

Re: Linux 2.2.18pre11

2000-09-28 Thread Marcelo Tosatti
On Fri, 29 Sep 2000, Eyal Lebedinsky wrote: Alan Cox wrote: 2.2.18pre11 I should mention that using an almost-all-modularised config I get this for the last few pachlevels: depmod: *** Unresolved symbols in /lib/modules/2.2.18pre11/misc/rio.o This patch fixes this: ---

Re: Kernal assert

2000-09-29 Thread Marcelo Tosatti
On Fri, 29 Sep 2000 [EMAIL PROTECTED] wrote: Can you assist? Sep 29 11:46:06 plato kernel: Unable to handle kernel paging request at virtual address 40ab06c8 Sep 29 11:46:06 plato kernel: current-tss.cr3 = 00101000, %cr3 = 00101000 Sep 29 11:46:06 plato kernel: *pde = Sep 29

Re: RTL8139 still doesn't work for me

2000-10-02 Thread Marcelo Tosatti
On Tue, 3 Oct 2000, Simon Richter wrote: Hi, I'm running 2.2.17 with the rtl8139 fix from 2.2.18pre, and after about two hours of normal operation (no crashes, no fs corruption -- Thanks Jeff) the network suddenly stops responding. Calling "ifconfig" (just looking at the stats) sometimes

Re: fun ?

2000-09-11 Thread Marcelo Tosatti
On Mon, 11 Sep 2000, Marcelo Tosatti wrote: On Mon, 11 Sep 2000, octave klaba wrote: Hello, upgrading from 2.2.16 to 2.2.17 a raid-soft config (adaptec 5x36Go) /sbin/lilo gave a D process ( :) ) root 14823 0.0 0.1 1184 496 ?DSep10 0:00 /sbin/lilo

Re: fun ?

2000-09-11 Thread Marcelo Tosatti
On Mon, 11 Sep 2000, octave klaba wrote: Hello, upgrading from 2.2.16 to 2.2.17 a raid-soft config (adaptec 5x36Go) /sbin/lilo gave a D process ( :) ) root 14823 0.0 0.1 1184 496 ?DSep10 0:00 /sbin/lilo one question: reboot or not to reboot ? I have no flopy

Re: BUG in tcp.c ?

2000-10-06 Thread Marcelo Tosatti
On Fri, 6 Oct 2000, David S. Miller wrote: Date: Fri, 6 Oct 2000 19:25:38 -0300 (BRST) From: Rik van Riel [EMAIL PROTECTED] Is this an actual bug, or am I overlooking something? It is a bug and I'll change TCP's sendmsg to use sk-allocation as it should. Thanks for pointing

__bad_udelay in 2.2.18pre15

2000-10-10 Thread Marcelo Tosatti
2.2.18pre15 defines udelay as (in file include/asm-i386/delay.h) : ... extern void __bad_udelay(void); ... #define udelay(n) (__builtin_constant_p(n) ? \ ((n) 2 ? __bad_udelay() : __const_udelay((n) * 0x10c6ul)) : \ __udelay(n)) ... It seems __bad_udelay is not

Re: 2.2.17 Crash

2000-10-11 Thread Marcelo Tosatti
On Wed, 11 Oct 2000, Mike Elmore wrote: All, Had a crash this morning for the first time in a while... 2.2.17 Locked up Cold. Machine is a SMP 2xPII450 w/ 128M RAM on a Tyan Tiger100 BX board. Here's the kernel output: Oct 11 08:29:30

Re: 2.2.17 Crash

2000-10-12 Thread Marcelo Tosatti
On Thu, 12 Oct 2000, Andrea Arcangeli wrote: On Wed, Oct 11, 2000 at 03:35:40PM -0200, Marcelo Tosatti wrote: Now I'm not sure if this can be caused by a memory problem. It can. Ok, thanks. Mike, could you try to run memtest86 (you can find it at freshmeat) to find out if your

Re: Linux 2.2.18pre17

2000-10-19 Thread Marcelo Tosatti
On Thu, 19 Oct 2000, Alan Cox wrote: This is just to give folks something to sync against. Test it by all means however. Must fix stuff left to do for 2.2.18final - Merge the S/390 stuff and make S/390 build again - Fix the megaraid (revert if need be) - Fix the ps/2

Re: Linux 2.2.18pre17

2000-10-20 Thread Marcelo Tosatti
On Thu, 19 Oct 2000, Alan Cox wrote: - Get to the bottom of the VM mystery if possible The RAID problem (which is caused by VM changes) is the same deadlock found in drbd and nbd. It was not a problem with kernels 2.2.17 because there was no write throttling in shrink_mmap. I'm

Re: 2.2.17 kernel panic

2000-10-25 Thread Marcelo Tosatti
Octave, Andrea fixed a corruption problem which looks exactly what you're hitting. Please try ftp://ftp.kernel.org/pub/people/andrea/patches/v2.2/2.2.18pre17/VM-global-2.2.18pre17-7.bz2 On Wed, 25 Oct 2000, octave klaba wrote: Hi, We test a smp server (bi-piii) and we have some problems

Re: [ANOMALIES]: 2.4.2 - __alloc_pages: failed - Patch failed

2001-02-24 Thread Marcelo Tosatti
On Fri, 23 Feb 2001, Shawn Starr wrote: Feb 23 21:17:47 coredump kernel: __alloc_pages: 3-order allocation failed. Feb 23 21:17:47 coredump kernel: __alloc_pages: 2-order allocation failed. Feb 23 21:17:47 coredump kernel: __alloc_pages: 1-order allocation failed. Feb 23 21:17:47 coredump

Re: [ANOMALIES]: 2.4.2 - __alloc_pages: failed - Patch failed

2001-02-25 Thread Marcelo Tosatti
On Sun, 25 Feb 2001, Mike Galbraith wrote: The way sg_low_malloc() tries to allocate, failure messages are pretty much garanteed. It tries high order allocations (which are unreliable even when not stressed) and backs off until it succeeds. In other words, the messages are a red

  1   2   3   4   5   6   7   8   9   10   >