From owner-linux-mm-bugs@nl.linux.org Thu Mar 1 23:54:00 2001 Received: by humbolt.nl.linux.org id ; Thu, 1 Mar 2001 23:53:49 +0100 Received: from distro.conectiva.com.br ([200.250.58.149]:27660 "EHLO distro.conectiva.com.br") by humbolt.nl.linux.org with ESMTP id ; Thu, 1 Mar 2001 23:53:36 +0100 Received: (from nobody@localhost) by distro.conectiva.com.br (8.9.3/8.9.3) id TAA00573; Thu, 1 Mar 2001 19:51:54 -0300 Date: Thu, 1 Mar 2001 19:51:54 -0300 Message-Id: <200103012251.TAA00573@distro.conectiva.com.br> From: bugzilla-daemon@distro.conectiva.com.br To: linux-mm-bugs@nl.linux.org Subject: [Bug 1186 (linux mm)] Changed - 2.4.1: OOM killer doesn't trigger Sender: owner-linux-mm-bugs@nl.linux.org Precedence: bulk Reply-To: linux-mm-bugs@nl.linux.org Return-Path: X-Orcpt: rfc822;linux-mm-bugs-list http://distro.conectiva.com.br/bugzilla/show_bug.cgi?id=1186 ------- Additional Comments From riel@conectiva.com.br 2001-03-01 19:51 ------- *** Bug 1594 has been marked as a duplicate of this bug. *** - Linux-mm-bugs: bugzilla list for the Linux-MM subsystem Archive: http://mail.nl.linux.org/linux-mm-bugs/ Web site: http://www.linux.eu.org/Linux-MM/ Development: linux-mm@kvack.org From owner-linux-mm-bugs@nl.linux.org Thu Mar 1 23:54:45 2001 Received: by humbolt.nl.linux.org id ; Thu, 1 Mar 2001 23:54:20 +0100 Received: from distro.conectiva.com.br ([200.250.58.149]:28172 "EHLO distro.conectiva.com.br") by humbolt.nl.linux.org with ESMTP id ; Thu, 1 Mar 2001 23:53:46 +0100 Received: (from nobody@localhost) by distro.conectiva.com.br (8.9.3/8.9.3) id TAA00482; Thu, 1 Mar 2001 19:47:01 -0300 Date: Thu, 1 Mar 2001 19:47:01 -0300 Message-Id: <200103012247.TAA00482@distro.conectiva.com.br> From: bugzilla-daemon@distro.conectiva.com.br To: linux-mm-bugs@nl.linux.org Subject: [Bug 1186 (linux mm)] Changed - 2.4.1: OOM killer doesn't trigger Sender: owner-linux-mm-bugs@nl.linux.org Precedence: bulk Reply-To: linux-mm-bugs@nl.linux.org Return-Path: X-Orcpt: rfc822;linux-mm-bugs-list http://distro.conectiva.com.br/bugzilla/show_bug.cgi?id=1186 ------- Additional Comments From riel@conectiva.com.br 2001-03-01 19:46 ------- Created an attachment (id=52) patch to make the OOM killer trigger again (please test) - Linux-mm-bugs: bugzilla list for the Linux-MM subsystem Archive: http://mail.nl.linux.org/linux-mm-bugs/ Web site: http://www.linux.eu.org/Linux-MM/ Development: linux-mm@kvack.org From owner-linux-mm-bugs@nl.linux.org Thu Mar 1 23:55:20 2001 Received: by humbolt.nl.linux.org id ; Thu, 1 Mar 2001 23:54:57 +0100 Received: from distro.conectiva.com.br ([200.250.58.149]:28940 "EHLO distro.conectiva.com.br") by humbolt.nl.linux.org with ESMTP id ; Thu, 1 Mar 2001 23:53:56 +0100 Received: (from nobody@localhost) by distro.conectiva.com.br (8.9.3/8.9.3) id TAA00579; Thu, 1 Mar 2001 19:51:55 -0300 Date: Thu, 1 Mar 2001 19:51:55 -0300 Message-Id: <200103012251.TAA00579@distro.conectiva.com.br> From: bugzilla-daemon@distro.conectiva.com.br To: linux-mm-bugs@nl.linux.org Subject: [Bug 1594 (linux mm)] Changed - mm bug Sender: owner-linux-mm-bugs@nl.linux.org Precedence: bulk Reply-To: linux-mm-bugs@nl.linux.org Return-Path: X-Orcpt: rfc822;linux-mm-bugs-list http://distro.conectiva.com.br/bugzilla/show_bug.cgi?id=1594 riel@conectiva.com.br changed: What |Old Value |New Value ---------------------------------------------------------------------------- Status|UNCONFIRMED |RESOLVED Resolution| |DUPLICATE ------- Additional Comments From riel@conectiva.com.br 2001-03-01 19:51 ------- ok, after some discussion with Xose this turned out to be a case of the OOM killer not being run (while nr_free_pages() == freepages.min). This is a duplicate of bug 1186 and should be fixed by the patch for that bug ... please test. *** This bug has been marked as a duplicate of 1186 *** - Linux-mm-bugs: bugzilla list for the Linux-MM subsystem Archive: http://mail.nl.linux.org/linux-mm-bugs/ Web site: http://www.linux.eu.org/Linux-MM/ Development: linux-mm@kvack.org From owner-linux-mm-bugs@nl.linux.org Fri Mar 2 00:04:04 2001 Received: by humbolt.nl.linux.org id ; Fri, 2 Mar 2001 00:03:54 +0100 Received: from distro.conectiva.com.br ([200.250.58.149]:15629 "EHLO distro.conectiva.com.br") by humbolt.nl.linux.org with ESMTP id ; Fri, 2 Mar 2001 00:03:33 +0100 Received: (from nobody@localhost) by distro.conectiva.com.br (8.9.3/8.9.3) id TAA00663; Thu, 1 Mar 2001 19:54:34 -0300 Date: Thu, 1 Mar 2001 19:54:34 -0300 Message-Id: <200103012254.TAA00663@distro.conectiva.com.br> From: bugzilla-daemon@distro.conectiva.com.br To: linux-mm-bugs@nl.linux.org Subject: [Bug 1403 (linux mm)] Changed - stalls with tmpfs Sender: owner-linux-mm-bugs@nl.linux.org Precedence: bulk Reply-To: linux-mm-bugs@nl.linux.org Return-Path: X-Orcpt: rfc822;linux-mm-bugs-list http://distro.conectiva.com.br/bugzilla/show_bug.cgi?id=1403 ------- Additional Comments From riel@conectiva.com.br 2001-03-01 19:54 ------- arjan, could you please test the page_launder() patches in the -ac kernels to see if the bug still exists ? - Linux-mm-bugs: bugzilla list for the Linux-MM subsystem Archive: http://mail.nl.linux.org/linux-mm-bugs/ Web site: http://www.linux.eu.org/Linux-MM/ Development: linux-mm@kvack.org From owner-linux-mm-bugs@nl.linux.org Fri Mar 2 17:54:09 2001 Received: by humbolt.nl.linux.org id ; Fri, 2 Mar 2001 17:53:57 +0100 Received: from distro.conectiva.com.br ([200.250.58.149]:5128 "EHLO distro.conectiva.com.br") by humbolt.nl.linux.org with ESMTP id ; Fri, 2 Mar 2001 17:53:35 +0100 Received: (from nobody@localhost) by distro.conectiva.com.br (8.9.3/8.9.3) id NAA16337; Fri, 2 Mar 2001 13:42:28 -0300 Date: Fri, 2 Mar 2001 13:42:28 -0300 Message-Id: <200103021642.NAA16337@distro.conectiva.com.br> From: bugzilla-daemon@distro.conectiva.com.br To: linux-mm-bugs@nl.linux.org Subject: [Bug 1622 (linux mm)] Changed - Incoherent Count of Users of Task Substructures and Direct Memory Access Bugs in 2.4.2 Sender: owner-linux-mm-bugs@nl.linux.org Precedence: bulk Reply-To: linux-mm-bugs@nl.linux.org Return-Path: X-Orcpt: rfc822;linux-mm-bugs-list http://distro.conectiva.com.br/bugzilla/show_bug.cgi?id=1622 riel@conectiva.com.br changed: What |Old Value |New Value ---------------------------------------------------------------------------- Status|UNCONFIRMED |RESOLVED Resolution| |FIXED ------- Additional Comments From riel@conectiva.com.br 2001-03-02 13:42 ------- the patches look sane to me, but since they're mostly outside of the VM area of the kernel, we should probably discuss them with Linus and co. Moshe, could you send these (+comments) to linux-kernel@vger.kernel.org, please ? - Linux-mm-bugs: bugzilla list for the Linux-MM subsystem Archive: http://mail.nl.linux.org/linux-mm-bugs/ Web site: http://www.linux.eu.org/Linux-MM/ Development: linux-mm@kvack.org From owner-linux-mm-bugs@nl.linux.org Fri Mar 2 22:04:01 2001 Received: by humbolt.nl.linux.org id ; Fri, 2 Mar 2001 22:03:52 +0100 Received: from distro.conectiva.com.br ([200.250.58.149]:10 "EHLO distro.conectiva.com.br") by humbolt.nl.linux.org with ESMTP id ; Fri, 2 Mar 2001 22:03:37 +0100 Received: (from nobody@localhost) by distro.conectiva.com.br (8.9.3/8.9.3) id RAA20893; Fri, 2 Mar 2001 17:56:34 -0300 Date: Fri, 2 Mar 2001 17:56:34 -0300 Message-Id: <200103022056.RAA20893@distro.conectiva.com.br> From: bugzilla-daemon@distro.conectiva.com.br To: linux-mm-bugs@nl.linux.org Subject: [Bug 1653 (linux mm)] New - port badram patch Sender: owner-linux-mm-bugs@nl.linux.org Precedence: bulk Reply-To: linux-mm-bugs@nl.linux.org Return-Path: X-Orcpt: rfc822;linux-mm-bugs-list http://distro.conectiva.com.br/bugzilla/show_bug.cgi?id=1653 Summary: port badram patch Product: Linux kernel MM Version: 2.4 (stable) Platform: All URL: OS/Version: Linux Status: NEW Status Whiteboard: Keywords: Resolution: Severity: enhancement Priority: P3 Component: linux mm AssignedTo: linux-mm-bugs@nl.linux.org ReportedBy: riel@conectiva.com.br after coming in the possession (sp?) of a 512MB DIMM, I suddenly have this urge to get Rick van Rein's badram patch ported to 2.4 and integrated into the kernel ... - Linux-mm-bugs: bugzilla list for the Linux-MM subsystem Archive: http://mail.nl.linux.org/linux-mm-bugs/ Web site: http://www.linux.eu.org/Linux-MM/ Development: linux-mm@kvack.org From owner-linux-mm-bugs@nl.linux.org Fri Mar 2 22:04:33 2001 Received: by humbolt.nl.linux.org id ; Fri, 2 Mar 2001 22:04:19 +0100 Received: from distro.conectiva.com.br ([200.250.58.149]:778 "EHLO distro.conectiva.com.br") by humbolt.nl.linux.org with ESMTP id ; Fri, 2 Mar 2001 22:03:48 +0100 Received: (from nobody@localhost) by distro.conectiva.com.br (8.9.3/8.9.3) id RAA20966; Fri, 2 Mar 2001 17:59:25 -0300 Date: Fri, 2 Mar 2001 17:59:25 -0300 Message-Id: <200103022059.RAA20966@distro.conectiva.com.br> From: bugzilla-daemon@distro.conectiva.com.br To: linux-mm-bugs@nl.linux.org Subject: [Bug 1654 (linux mm)] New - ECC memory error detection + unmapping memory, etc... Sender: owner-linux-mm-bugs@nl.linux.org Precedence: bulk Reply-To: linux-mm-bugs@nl.linux.org Return-Path: X-Orcpt: rfc822;linux-mm-bugs-list http://distro.conectiva.com.br/bugzilla/show_bug.cgi?id=1654 Summary: ECC memory error detection + unmapping memory, etc... Product: Linux kernel MM Version: 2.4 (stable) Platform: All URL: OS/Version: Linux Status: NEW Status Whiteboard: Keywords: Resolution: Severity: enhancement Priority: P3 Component: linux mm AssignedTo: linux-mm-bugs@nl.linux.org ReportedBy: riel@conectiva.com.br in addition to the badram patch, it would be great if bluesmoke.c (or another file in the kernel source) would be enhanced to detect when the machine alerts the kernel to an ECC error in ram after that, the kernel could proceed to free up the memory, make sure it is never used again and (maybe) kill the task that was using that memory it should also print out a message telling the user which address was failing, so that we know which address to give to badram on the next boot ... - Linux-mm-bugs: bugzilla list for the Linux-MM subsystem Archive: http://mail.nl.linux.org/linux-mm-bugs/ Web site: http://www.linux.eu.org/Linux-MM/ Development: linux-mm@kvack.org From owner-linux-mm-bugs@nl.linux.org Sun Mar 4 21:44:21 2001 Received: by humbolt.nl.linux.org id ; Sun, 4 Mar 2001 21:44:02 +0100 Received: from distro.conectiva.com.br ([200.250.58.149]:58888 "EHLO distro.conectiva.com.br") by humbolt.nl.linux.org with ESMTP id ; Sun, 4 Mar 2001 21:43:31 +0100 Received: (from nobody@localhost) by distro.conectiva.com.br (8.9.3/8.9.3) id RAA17953; Sun, 4 Mar 2001 17:39:41 -0300 Date: Sun, 4 Mar 2001 17:39:41 -0300 Message-Id: <200103042039.RAA17953@distro.conectiva.com.br> From: bugzilla-daemon@distro.conectiva.com.br To: linux-mm-bugs@nl.linux.org Subject: [Bug 1657 (linux mm)] New - page table sharing for huge shared segments (mainly x86 ??) Sender: owner-linux-mm-bugs@nl.linux.org Precedence: bulk Reply-To: linux-mm-bugs@nl.linux.org Return-Path: X-Orcpt: rfc822;linux-mm-bugs-list http://distro.conectiva.com.br/bugzilla/show_bug.cgi?id=1657 Summary: page table sharing for huge shared segments (mainly x86 ??) Product: Linux kernel MM Version: 2.5 (beta) Platform: All URL: OS/Version: Linux Status: UNCONFIRMED Status Whiteboard: Keywords: Resolution: Severity: enhancement Priority: P3 Component: linux mm AssignedTo: linux-mm-bugs@nl.linux.org ReportedBy: riel@conectiva.com.br On an Oracle / SAP workload Linux has the following problems: 1. 850 MB limit on kernel direct mapped memory 2. suppose 500 processes w/ 1GB shared segment -> 500MB of pagetable overhead wastes memory and fills up kernel virtual address space 3. evicting pages requires us to scan 500 address spaces to swap out the page If, however, we force the large mappings of SHM segments and files to be 4MB aligned, we can share the page tables and: 1. reduce page table overhead 2. don't fill up kernel virtual memory 3. make page eviction easier An easy way to achieve this would be to have an "own" mm_struct for the shared segments and simply "attach" the page tables of the shared segments into the tasks. That way kswapd can scan the page tables of the shared segments to evict pages and overhead will just be lower... Comments ? - Linux-mm-bugs: bugzilla list for the Linux-MM subsystem Archive: http://mail.nl.linux.org/linux-mm-bugs/ Web site: http://www.linux.eu.org/Linux-MM/ Development: linux-mm@kvack.org From owner-linux-mm-bugs@nl.linux.org Sun Mar 11 03:03:57 2001 Received: by humbolt.nl.linux.org id ; Sun, 11 Mar 2001 03:03:47 +0100 Received: from distro.conectiva.com.br ([200.250.58.149]:33294 "EHLO distro.conectiva.com.br") by humbolt.nl.linux.org with ESMTP id ; Sun, 11 Mar 2001 03:03:20 +0100 Received: (from nobody@localhost) by distro.conectiva.com.br (8.9.3/8.9.3) id WAA10754; Sat, 10 Mar 2001 22:59:30 -0300 Date: Sat, 10 Mar 2001 22:59:30 -0300 Message-Id: <200103110159.WAA10754@distro.conectiva.com.br> From: bugzilla-daemon@distro.conectiva.com.br To: linux-mm-bugs@nl.linux.org Subject: [Bug 1566 (linux mm)] Changed - vmbalance patch in ac7 triggers OOM killer too often Sender: owner-linux-mm-bugs@nl.linux.org Precedence: bulk Reply-To: linux-mm-bugs@nl.linux.org Return-Path: X-Orcpt: rfc822;linux-mm-bugs-list http://distro.conectiva.com.br/bugzilla/show_bug.cgi?id=1566 riel@conectiva.com.br changed: What |Old Value |New Value ---------------------------------------------------------------------------- Status|UNCONFIRMED |RESOLVED Resolution| |LATER ------- Additional Comments From riel@conectiva.com.br 2001-03-10 22:59 ------- reclaiming swap space on swapin (or refill_inactive_scan??) is on the TODO list ... - Linux-mm-bugs: bugzilla list for the Linux-MM subsystem Archive: http://mail.nl.linux.org/linux-mm-bugs/ Web site: http://www.linux.eu.org/Linux-MM/ Development: linux-mm@kvack.org From owner-linux-mm-bugs@nl.linux.org Tue Mar 13 00:33:42 2001 Received: by humbolt.nl.linux.org id ; Tue, 13 Mar 2001 00:33:29 +0100 Received: from distro.conectiva.com.br ([200.250.58.149]:48141 "EHLO distro.conectiva.com.br") by humbolt.nl.linux.org with ESMTP id ; Tue, 13 Mar 2001 00:33:11 +0100 Received: (from nobody@localhost) by distro.conectiva.com.br (8.9.3/8.9.3) id UAA14766; Mon, 12 Mar 2001 20:24:48 -0300 Date: Mon, 12 Mar 2001 20:24:48 -0300 Message-Id: <200103122324.UAA14766@distro.conectiva.com.br> From: bugzilla-daemon@distro.conectiva.com.br To: linux-mm-bugs@nl.linux.org Subject: [Bug 1654 (linux mm)] Changed - ECC memory error detection + unmapping memory, etc... Sender: owner-linux-mm-bugs@nl.linux.org Precedence: bulk Reply-To: linux-mm-bugs@nl.linux.org Return-Path: X-Orcpt: rfc822;linux-mm-bugs-list http://distro.conectiva.com.br/bugzilla/show_bug.cgi?id=1654 ------- Additional Comments From riel@conectiva.com.br 2001-03-12 20:24 ------- It seems there is some ECC supporting code ... http://www.anime.net/~goemon/linux-ecc/ I don't know in what state it is, however. - Linux-mm-bugs: bugzilla list for the Linux-MM subsystem Archive: http://mail.nl.linux.org/linux-mm-bugs/ Web site: http://www.linux.eu.org/Linux-MM/ Development: linux-mm@kvack.org From owner-linux-mm-bugs@nl.linux.org Tue Mar 13 01:53:56 2001 Received: by humbolt.nl.linux.org id ; Tue, 13 Mar 2001 01:53:35 +0100 Received: from distro.conectiva.com.br ([200.250.58.149]:16142 "EHLO distro.conectiva.com.br") by humbolt.nl.linux.org with ESMTP id ; Tue, 13 Mar 2001 01:53:13 +0100 Received: (from nobody@localhost) by distro.conectiva.com.br (8.9.3/8.9.3) id VAA15742; Mon, 12 Mar 2001 21:44:04 -0300 Date: Mon, 12 Mar 2001 21:44:04 -0300 Message-Id: <200103130044.VAA15742@distro.conectiva.com.br> From: bugzilla-daemon@distro.conectiva.com.br To: linux-mm-bugs@nl.linux.org Subject: [Bug 1203 (linux mm)] Changed - 2.5: load control Sender: owner-linux-mm-bugs@nl.linux.org Precedence: bulk Reply-To: linux-mm-bugs@nl.linux.org Return-Path: X-Orcpt: rfc822;linux-mm-bugs-list http://distro.conectiva.com.br/bugzilla/show_bug.cgi?id=1203 riel@conectiva.com.br changed: What |Old Value |New Value ---------------------------------------------------------------------------- Status|NEW |ASSIGNED OS/Version| |Linux Platform| |All ------- Additional Comments From riel@conectiva.com.br 2001-03-12 21:44 ------- OK, this thing seems to be absolutely trivial to do and also quite necessary ... I've seen another few Linux machines thrash to a halt recently (under test loads and under heavy overload). I guess I'll give this thing a go... - Linux-mm-bugs: bugzilla list for the Linux-MM subsystem Archive: http://mail.nl.linux.org/linux-mm-bugs/ Web site: http://www.linux.eu.org/Linux-MM/ Development: linux-mm@kvack.org From owner-linux-mm-bugs@nl.linux.org Thu Mar 15 14:03:47 2001 Received: by humbolt.nl.linux.org id ; Thu, 15 Mar 2001 14:03:27 +0100 Received: from distro.conectiva.com.br ([200.250.58.149]:33797 "EHLO distro.conectiva.com.br") by humbolt.nl.linux.org with ESMTP id ; Thu, 15 Mar 2001 14:03:08 +0100 Received: (from nobody@localhost) by distro.conectiva.com.br (8.9.3/8.9.3) id JAA02693; Thu, 15 Mar 2001 09:59:59 -0300 Date: Thu, 15 Mar 2001 09:59:59 -0300 Message-Id: <200103151259.JAA02693@distro.conectiva.com.br> From: bugzilla-daemon@distro.conectiva.com.br To: linux-mm-bugs@nl.linux.org Subject: [Bug 1699 (linux mm)] New - change mmap_sem to R/W lock Sender: owner-linux-mm-bugs@nl.linux.org Precedence: bulk Reply-To: linux-mm-bugs@nl.linux.org Return-Path: X-Orcpt: rfc822;linux-mm-bugs-list http://distro.conectiva.com.br/bugzilla/show_bug.cgi?id=1699 Summary: change mmap_sem to R/W lock Product: Linux kernel MM Version: 2.4 (stable) Platform: All URL: OS/Version: Linux Status: NEW Status Whiteboard: Keywords: Resolution: Severity: enhancement Priority: P3 Component: linux mm AssignedTo: linux-mm-bugs@nl.linux.org ReportedBy: riel@conectiva.com.br it happens too often that page faults in other programs keep vmstat and top from working because they hold the mm->mmap_sem for a long time this is stupid because both the pagefault path and procfs use the mm->mmap_sem as a READ LOCK, to prevent the VMA list from changing under them changing the mm->mmap_sem to a read/write semaphore would fix this problem it would, however, need us to introduce another lock in the page fault path to make sure we can't have multiple parallel pagefaults on the same address (range?) ... having one per-mm pagefault lock should fix this trivially, however - Linux-mm-bugs: bugzilla list for the Linux-MM subsystem Archive: http://mail.nl.linux.org/linux-mm-bugs/ Web site: http://www.linux.eu.org/Linux-MM/ Development: linux-mm@kvack.org From owner-linux-mm-bugs@nl.linux.org Thu Mar 15 14:43:42 2001 Received: by humbolt.nl.linux.org id ; Thu, 15 Mar 2001 14:43:31 +0100 Received: from distro.conectiva.com.br ([200.250.58.149]:10502 "EHLO distro.conectiva.com.br") by humbolt.nl.linux.org with ESMTP id ; Thu, 15 Mar 2001 14:43:06 +0100 Received: (from nobody@localhost) by distro.conectiva.com.br (8.9.3/8.9.3) id KAA03472; Thu, 15 Mar 2001 10:40:02 -0300 Date: Thu, 15 Mar 2001 10:40:02 -0300 Message-Id: <200103151340.KAA03472@distro.conectiva.com.br> From: bugzilla-daemon@distro.conectiva.com.br To: linux-mm-bugs@nl.linux.org Subject: [Bug 1699 (linux mm)] Changed - change mmap_sem to R/W lock Sender: owner-linux-mm-bugs@nl.linux.org Precedence: bulk Reply-To: linux-mm-bugs@nl.linux.org Return-Path: X-Orcpt: rfc822;linux-mm-bugs-list http://distro.conectiva.com.br/bugzilla/show_bug.cgi?id=1699 ------- Additional Comments From sct@redhat.com 2001-03-15 10:40 ------- We already have a per-mm page fault lock, the page_table_lock spinlock. It's pretty trivial to avoid page fault races with that existing lock. The worst you have to do is to make sure that you do a consistent read/modify/write of the pte without dropping the lock, so you have to be prepared to return early from the page fault and try again if you detect that a fault has already happened by the time you get around to taking or retaking the lock yourself. It's a 2.5 issue --- there are too many places for subtle things to go wrong --- but it's not a major rearchitecture. - Linux-mm-bugs: bugzilla list for the Linux-MM subsystem Archive: http://mail.nl.linux.org/linux-mm-bugs/ Web site: http://www.linux.eu.org/Linux-MM/ Development: linux-mm@kvack.org From owner-linux-mm-bugs@nl.linux.org Sat Mar 17 12:03:35 2001 Received: by humbolt.nl.linux.org id ; Sat, 17 Mar 2001 12:03:23 +0100 Received: from distro.conectiva.com.br ([200.250.58.149]:8460 "EHLO distro.conectiva.com.br") by humbolt.nl.linux.org with ESMTP id ; Sat, 17 Mar 2001 12:02:58 +0100 Received: (from nobody@localhost) by distro.conectiva.com.br (8.9.3/8.9.3) id HAA11280; Sat, 17 Mar 2001 07:59:19 -0300 Date: Sat, 17 Mar 2001 07:59:19 -0300 Message-Id: <200103171059.HAA11280@distro.conectiva.com.br> From: bugzilla-daemon@distro.conectiva.com.br To: linux-mm-bugs@nl.linux.org Subject: [Bug 1653 (linux mm)] Changed - port badram patch Sender: owner-linux-mm-bugs@nl.linux.org Precedence: bulk Reply-To: linux-mm-bugs@nl.linux.org Return-Path: X-Orcpt: rfc822;linux-mm-bugs-list http://distro.conectiva.com.br/bugzilla/show_bug.cgi?id=1653 ------- Additional Comments From ee71ae@eim.surrey.ac.uk 2001-03-17 07:59 ------- This has already been done by Nico Schmoigl @ nico@writemail.com See http://webrum.uni-mannheim.de/math/schmoigl/linux/ Andrew (aka jfreak) - Linux-mm-bugs: bugzilla list for the Linux-MM subsystem Archive: http://mail.nl.linux.org/linux-mm-bugs/ Web site: http://www.linux.eu.org/Linux-MM/ Development: linux-mm@kvack.org From owner-linux-mm-bugs@nl.linux.org Wed Mar 21 04:43:24 2001 Received: by humbolt.nl.linux.org id ; Wed, 21 Mar 2001 04:43:13 +0100 Received: from distro.conectiva.com.br ([200.250.58.149]:39434 "EHLO distro.conectiva.com.br") by humbolt.nl.linux.org with ESMTP id ; Wed, 21 Mar 2001 04:42:56 +0100 Received: (from nobody@localhost) by distro.conectiva.com.br (8.9.3/8.9.3) id AAA25325; Wed, 21 Mar 2001 00:35:00 -0300 Date: Wed, 21 Mar 2001 00:35:00 -0300 Message-Id: <200103210335.AAA25325@distro.conectiva.com.br> From: bugzilla-daemon@distro.conectiva.com.br To: linux-mm-bugs@nl.linux.org Subject: [Bug 1699 (linux mm)] Changed - change mmap_sem to R/W lock Sender: owner-linux-mm-bugs@nl.linux.org Precedence: bulk Reply-To: linux-mm-bugs@nl.linux.org Return-Path: X-Orcpt: rfc822;linux-mm-bugs-list http://distro.conectiva.com.br/bugzilla/show_bug.cgi?id=1699 riel@conectiva.com.br changed: What |Old Value |New Value ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |FIXED ------- Additional Comments From riel@conectiva.com.br 2001-03-21 00:34 ------- merged with Linus' tree ... it turned out to be quite trivial ;) now lets see exactly how much performance of MySQL, apache2 and other threaded apps will improve ... - Linux-mm-bugs: bugzilla list for the Linux-MM subsystem Archive: http://mail.nl.linux.org/linux-mm-bugs/ Web site: http://www.linux.eu.org/Linux-MM/ Development: linux-mm@kvack.org From owner-linux-mm-bugs@nl.linux.org Wed Mar 21 21:13:54 2001 Received: by humbolt.nl.linux.org id ; Wed, 21 Mar 2001 21:13:39 +0100 Received: from distro.conectiva.com.br ([200.250.58.149]:47114 "EHLO distro.conectiva.com.br") by humbolt.nl.linux.org with ESMTP id ; Wed, 21 Mar 2001 21:13:00 +0100 Received: (from nobody@localhost) by distro.conectiva.com.br (8.9.3/8.9.3) id RAA09769; Wed, 21 Mar 2001 17:07:58 -0300 Date: Wed, 21 Mar 2001 17:07:58 -0300 Message-Id: <200103212007.RAA09769@distro.conectiva.com.br> From: bugzilla-daemon@distro.conectiva.com.br To: linux-mm-bugs@nl.linux.org Subject: [Bug 1186 (linux mm)] Changed - 2.4.1: OOM killer doesn't trigger Sender: owner-linux-mm-bugs@nl.linux.org Precedence: bulk Reply-To: linux-mm-bugs@nl.linux.org Return-Path: X-Orcpt: rfc822;linux-mm-bugs-list http://distro.conectiva.com.br/bugzilla/show_bug.cgi?id=1186 ------- Additional Comments From ingo.oeser@informatik.tu-chemnitz.de 2001-03-21 17:07 ------- In ac19 and ac20 this Bug does still trigger. Scenario: A Cyrix III 500 with 64MB cramfs over loop file of 10MB size 3x ramfs with limit of 16M (but not used too much) 16MB ext2 / (which also holds the loop file) 20G of other ext2 partitions System runs iptables over ipchains module doing forwarding and filtering. Also running dhcpd and an eth0 (8139C, 3c590C-TX both the same results) and lo. script to trigger: nc localhost 7 /dev/null & nc localhost 7 /dev/null & nc localhost 7 /dev/null & nc localhost 7 /dev/null & vmstat 1 And watch it die in some seconds. Also tested with ftp to this machine and smb to this machine. Adding another 64MB of RAM solved all problems. Using only 10Mbit-Networking also solved it. So it only shows up with 100MBit connections and maybe higher. I saw around up to 5000 cs and usally 3000 cs in vmstat with only 2000 ints. Adding the 64MB of RAM showed only some hundreds of cs. Can provide even more details on demand. - Linux-mm-bugs: bugzilla list for the Linux-MM subsystem Archive: http://mail.nl.linux.org/linux-mm-bugs/ Web site: http://www.linux.eu.org/Linux-MM/ Development: linux-mm@kvack.org From owner-linux-mm-bugs@nl.linux.org Wed Mar 28 20:46:56 2001 Received: by humbolt.nl.linux.org id ; Wed, 28 Mar 2001 20:46:44 +0200 Received: from distro.conectiva.com.br ([200.250.58.149]:46596 "EHLO distro.conectiva.com.br") by humbolt.nl.linux.org with ESMTP id ; Wed, 28 Mar 2001 20:46:25 +0200 Received: (from nobody@localhost) by distro.conectiva.com.br (8.9.3/8.9.3) id PAA21628; Wed, 28 Mar 2001 15:45:06 -0300 Date: Wed, 28 Mar 2001 15:45:06 -0300 Message-Id: <200103281845.PAA21628@distro.conectiva.com.br> From: bugzilla-daemon@distro.conectiva.com.br To: linux-mm-bugs@nl.linux.org Subject: [Bug 1186 (linux mm)] Changed - 2.4.1: OOM killer doesn't trigger Sender: owner-linux-mm-bugs@nl.linux.org Precedence: bulk Reply-To: linux-mm-bugs@nl.linux.org Return-Path: X-Orcpt: rfc822;linux-mm-bugs-list http://distro.conectiva.com.br/bugzilla/show_bug.cgi?id=1186 ------- Additional Comments From riel@conectiva.com.br 2001-03-28 15:45 ------- ok, so dledford found a case where the OOM killer triggers *too early* and we still have enough free swap ... this needs to be fixed and it appears people are working on it ... - Linux-mm-bugs: bugzilla list for the Linux-MM subsystem Archive: http://mail.nl.linux.org/linux-mm-bugs/ Web site: http://www.linux.eu.org/Linux-MM/ Development: linux-mm@kvack.org