From: Nikita Danilov <nikita@clusterfs.com> To: Andrew Morton <akpm@osdl.org> Cc: pmarques@grupopie.com, linux-mm@kvack.org, linux-kernel@vger.kernel.org, hch@infradead.org Subject: Re: [RFC] per thread page reservation patch Date: Sat, 08 Jan 2005 02:17:31 +0300 [thread overview] Message-ID: <m18y74rfqs.fsf@clusterfs.com> (raw) In-Reply-To: <20050107150315.3c1714a4.akpm@osdl.org> (Andrew Morton's message of "Fri, 7 Jan 2005 15:03:15 -0800") Andrew Morton <akpm@osdl.org> writes: > Nikita Danilov <nikita@clusterfs.com> wrote: >> >> > And the whole idea is pretty flaky really - how can one precalculate how >> > much memory an arbitrary md-on-dm-on-loop-on-md-on-NBD stack will want to >> > use? It really would be better if we could drop the whole patch and make >> > reiser4 behave more sanely when its writepage is called with for_reclaim=1. >> >> Reiser4 doesn't use this for ->writepage(), by the way. This is used by >> tree balancing code to assure that balancing cannot get -ENOMEM in the >> middle of tree modification, because undo is _so_ very complicated. > > Oh. And that involves performing I/O, yes? Yes, balancing may read tree or bitmap nodes from the disk. > > Why does the filesystem risk going oom during the rebalance anyway? Is it > doing atomic allocations? No, just __alloc_pages(GFP_KERNEL, 0, ...) returns NULL. When this happens, the only thing balancing can do is to panic. Nikita.
WARNING: multiple messages have this Message-ID (diff)
From: Nikita Danilov <nikita@clusterfs.com> To: Andrew Morton <akpm@osdl.org> Cc: pmarques@grupopie.com, linux-mm@kvack.org, linux-kernel@vger.kernel.org, hch@infradead.org Subject: Re: [RFC] per thread page reservation patch Date: Sat, 08 Jan 2005 02:17:31 +0300 [thread overview] Message-ID: <m18y74rfqs.fsf@clusterfs.com> (raw) In-Reply-To: <20050107150315.3c1714a4.akpm@osdl.org> (Andrew Morton's message of "Fri, 7 Jan 2005 15:03:15 -0800") Andrew Morton <akpm@osdl.org> writes: > Nikita Danilov <nikita@clusterfs.com> wrote: >> >> > And the whole idea is pretty flaky really - how can one precalculate how >> > much memory an arbitrary md-on-dm-on-loop-on-md-on-NBD stack will want to >> > use? It really would be better if we could drop the whole patch and make >> > reiser4 behave more sanely when its writepage is called with for_reclaim=1. >> >> Reiser4 doesn't use this for ->writepage(), by the way. This is used by >> tree balancing code to assure that balancing cannot get -ENOMEM in the >> middle of tree modification, because undo is _so_ very complicated. > > Oh. And that involves performing I/O, yes? Yes, balancing may read tree or bitmap nodes from the disk. > > Why does the filesystem risk going oom during the rebalance anyway? Is it > doing atomic allocations? No, just __alloc_pages(GFP_KERNEL, 0, ...) returns NULL. When this happens, the only thing balancing can do is to panic. Nikita. -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@kvack.org. For more info on Linux MM, see: http://www.linux-mm.org/ . Don't email: <a href=mailto:"aart@kvack.org"> aart@kvack.org </a>
next prev parent reply other threads:[~2005-01-07 23:21 UTC|newest] Thread overview: 286+ messages / expand[flat|nested] mbox.gz Atom feed top 2004-12-15 6:56 unregister_ioctl32_conversion and modules. ioctl32 revisited Andi Kleen 2004-12-15 7:46 ` Michael S. Tsirkin 2004-12-15 8:00 ` Andi Kleen 2004-12-15 8:21 ` Michael S. Tsirkin 2004-12-15 8:29 ` Andi Kleen 2004-12-15 11:42 ` Michael S. Tsirkin 2004-12-15 13:46 ` Arnd Bergmann 2004-12-15 16:12 ` Andi Kleen 2004-12-15 16:45 ` Arnd Bergmann 2004-12-15 16:57 ` Andi Kleen 2004-12-15 17:47 ` Arnd Bergmann 2004-12-15 17:59 ` Andi Kleen 2004-12-15 18:21 ` Michael S. Tsirkin 2004-12-16 4:06 ` Andi Kleen 2004-12-26 22:26 ` Chris Wedgwood 2004-12-26 22:49 ` [discuss] " Arnd Bergmann 2004-12-27 11:49 ` [discuss] Re: unregister_ioctl32_conversion and modules. ioct l32 revisited Michael S. Tsirkin 2005-01-05 15:25 ` Michael S. Tsirkin 2005-01-05 16:07 ` Arnd Bergmann 2004-12-15 18:20 ` unregister_ioctl32_conversion and modules. ioctl32 revisited Takashi Iwai 2004-12-15 18:30 ` Lee Revell 2004-12-15 19:34 ` Michael S. Tsirkin 2004-12-15 19:34 ` Michael S. Tsirkin 2004-12-16 5:03 ` [discuss] " Andi Kleen 2004-12-16 5:03 ` Andi Kleen 2004-12-16 7:53 ` Ingo Molnar 2004-12-16 8:09 ` Andi Kleen 2004-12-16 8:25 ` Andrew Morton 2004-12-16 8:25 ` Andrew Morton 2004-12-16 8:30 ` Michael S. Tsirkin 2004-12-16 8:38 ` Andi Kleen 2004-12-17 1:43 ` [PATCH] " Michael S. Tsirkin 2004-12-16 16:08 ` Christoph Hellwig 2005-01-03 9:11 ` 2.6.10-mm1 Andrew Morton 2005-01-03 10:07 ` 2.6.10-mm1 Christoph Hellwig 2005-01-03 17:19 ` 2.6.10-mm1 Jesse Barnes 2005-01-03 23:29 ` 2.6.10-mm1 [failure on AMD64] Rafael J. Wysocki 2005-01-03 23:30 ` Jesse Barnes 2005-01-04 11:23 ` Thomas Molina 2005-01-04 13:22 ` Rafael J. Wysocki 2005-01-04 12:22 ` Andi Kleen 2005-01-04 21:05 ` Rafael J. Wysocki 2005-01-06 11:32 ` 2.6.10-mm1 Christoph Hellwig 2005-01-06 13:04 ` 2.6.10-mm1 David Howells 2005-01-06 13:06 ` 2.6.10-mm1 Christoph Hellwig 2005-01-03 10:25 ` 2.6.10-mm1 Christoph Hellwig 2005-01-03 13:21 ` 2.6.10-mm1 Christoph Hellwig 2005-01-03 13:35 ` 2.6.10-mm1 Michael S. Tsirkin 2005-01-03 11:48 ` 2.6.10-mm1 Christoph Hellwig 2005-01-05 17:27 ` 2.6.10-mm1 Hans Reiser 2005-01-06 13:52 ` 2.6.10-mm1 Vladimir Saveliev 2005-01-07 14:46 ` 2.6.10-mm1 Christoph Hellwig 2005-01-07 17:16 ` [RFC] per thread page reservation patch Vladimir Saveliev 2005-01-07 18:48 ` Andrew Morton 2005-01-07 18:48 ` Andrew Morton 2005-01-07 20:21 ` Nikita Danilov 2005-01-07 20:21 ` Nikita Danilov 2005-01-07 19:05 ` Christoph Hellwig 2005-01-07 19:05 ` Christoph Hellwig 2005-01-07 19:12 ` Christoph Hellwig 2005-01-07 19:12 ` Christoph Hellwig 2005-01-07 19:21 ` Robert Love 2005-01-07 19:21 ` Robert Love 2005-01-07 20:48 ` Nikita Danilov 2005-01-07 20:48 ` Nikita Danilov 2005-01-07 20:54 ` Christoph Hellwig 2005-01-07 20:54 ` Christoph Hellwig 2005-01-07 21:00 ` Nikita Danilov 2005-01-07 21:00 ` Nikita Danilov 2005-01-07 21:07 ` Christoph Hellwig 2005-01-07 21:07 ` Christoph Hellwig 2005-01-07 19:14 ` Paulo Marques 2005-01-07 19:14 ` Paulo Marques 2005-01-07 19:32 ` Christoph Hellwig 2005-01-07 19:32 ` Christoph Hellwig 2005-01-07 19:42 ` Andi Kleen 2005-01-07 19:42 ` Andi Kleen 2005-01-07 20:55 ` Nikita Danilov 2005-01-07 20:55 ` Nikita Danilov 2005-01-07 21:24 ` Andrew Morton 2005-01-07 21:24 ` Andrew Morton 2005-01-07 21:24 ` Andi Kleen 2005-01-07 21:24 ` Andi Kleen 2005-01-07 22:12 ` Nikita Danilov 2005-01-07 22:12 ` Nikita Danilov 2005-01-07 23:03 ` Andrew Morton 2005-01-07 23:03 ` Andrew Morton 2005-01-07 23:17 ` Nikita Danilov [this message] 2005-01-07 23:17 ` Nikita Danilov 2005-01-07 23:43 ` Andrew Morton 2005-01-07 23:43 ` Andrew Morton 2005-01-08 12:44 ` Nikita Danilov 2005-01-08 12:44 ` Nikita Danilov 2005-01-08 13:43 ` Hugh Dickins 2005-01-08 13:43 ` Hugh Dickins 2005-01-09 11:35 ` Marcelo Tosatti 2005-01-09 11:35 ` Marcelo Tosatti 2005-01-09 18:16 ` Nikita Danilov 2005-01-09 18:16 ` Nikita Danilov 2005-01-25 16:39 ` reiser4 core patches: [Was: [RFC] per thread page reservation patch] Vladimir Saveliev 2005-01-25 16:39 ` Vladimir Saveliev 2005-01-27 10:37 ` Adrian Bunk 2005-01-27 10:37 ` Adrian Bunk 2005-01-27 11:01 ` Christoph Hellwig 2005-01-27 11:01 ` Christoph Hellwig 2005-01-03 11:51 ` 2.6.10-mm1 Christoph Hellwig 2005-01-04 9:04 ` 2.6.10-mm1 Ingo Molnar 2005-01-04 9:26 ` 2.6.10-mm1 Christoph Hellwig 2005-01-04 9:33 ` 2.6.10-mm1 Ingo Molnar 2005-01-03 15:13 ` 2.6.10-mm1 William Lee Irwin III 2005-01-03 17:17 ` 2.6.10-mm1 Jesse Barnes 2005-01-05 22:38 ` 2.6.10-mm1 Matthew Dobson 2005-01-03 20:42 ` [PATCH] pktcdvd: make two functions static Peter Osterlund 2005-01-03 20:45 ` [PATCH] pktcdvd: grep-friendly function prototypes Peter Osterlund 2005-01-03 20:47 ` [PATCH] pktcdvd: Small documentation update Peter Osterlund 2005-01-03 20:53 ` [PATCH] isofs: Remove useless include Peter Osterlund 2005-01-03 20:58 ` [PATCH] synaptics: Remove unused struct member variable Peter Osterlund 2005-01-03 22:25 ` [PATCH] pktcdvd: make two functions static Bartlomiej Zolnierkiewicz 2005-01-04 9:08 ` 2.6.10-mm1 Ingo Molnar 2005-01-05 14:40 ` [PATCH] deprecate (un)register_ioctl32_conversion Michael S. Tsirkin 2005-01-05 14:46 ` Christoph Hellwig 2005-01-05 15:03 ` Michael S. Tsirkin 2005-01-05 15:11 ` Christoph Hellwig 2005-01-05 21:33 ` Andrew Morton 2005-01-06 14:41 ` Michael S. Tsirkin 2005-01-06 14:55 ` Christoph Hellwig 2005-01-06 15:22 ` Michael S. Tsirkin 2005-01-06 15:30 ` Christoph Hellwig 2005-01-06 15:56 ` Michael S. Tsirkin 2005-01-05 15:19 ` Andi Kleen 2005-01-05 15:55 ` Christoph Hellwig 2005-01-05 18:23 ` Takashi Iwai 2005-01-05 18:23 ` Takashi Iwai 2005-01-05 21:34 ` Andrew Morton 2005-01-06 8:22 ` 2.6.10-mm2 Andrew Morton 2005-01-06 8:39 ` 2.6.10-mm2 Nick Piggin 2005-01-06 9:11 ` 2.6.10-mm2 Eyal Lebedinsky 2005-01-06 9:24 ` 2.6.10-mm2 Andrew Morton 2005-01-06 9:27 ` 2.6.10-mm2 Christoph Hellwig 2005-01-06 10:17 ` 2.6.10-mm2 Juri Prokofjev 2005-01-06 14:54 ` 2.6.10-mm2 Dave Airlie 2005-01-06 10:17 ` [PATCH 2.6.10-mm2] m32r: build fix Hirokazu Takata 2005-01-06 10:47 ` 2.6.10-mm2 Christoph Hellwig 2005-01-06 11:19 ` 2.6.10-mm2 error: redefinition of `struct cfq_io_context' Helge Hafting 2005-01-06 14:21 ` Helge Hafting 2005-01-06 12:06 ` 2.6.10-mm2 Marcos D. Marado Torres 2005-01-11 14:08 ` acpi_power_off on 2.6.10-mm3 WAS: 2.6.10-mm2 Marcos D. Marado Torres 2005-01-06 12:48 ` 2.6.10-mm2 Sam Ravnborg 2005-01-06 18:15 ` 2.6.10-mm2 Felipe Alfaro Solana 2005-01-06 14:51 ` [PATCH] fget_light/fput_light for ioctls Michael S. Tsirkin 2005-01-06 16:18 ` [PATCH] fget_light/fput_light for ioctls (fixed) Michael S. Tsirkin 2005-01-06 15:00 ` [patch] 2.6.10-mm2: remove umsdos MAINTAINERS entry Adrian Bunk 2005-01-06 15:03 ` [patch] 2.6.10-mm2: fix MTD_BLOCK2MTD dependency Adrian Bunk 2005-01-06 15:03 ` Adrian Bunk 2005-01-08 20:18 ` Jörn Engel 2005-01-08 20:18 ` Jörn Engel 2005-01-06 17:11 ` 2.6.10-mm2 Alexander Gran 2005-01-06 17:55 ` 2.6.10-mm2 Diego Calleja 2005-01-06 18:17 ` 2.6.10-mm2 Marc Ballarin 2005-01-06 17:48 ` 2.6.10-mm2: swsusp regression Rafael J. Wysocki 2005-01-06 22:52 ` Pavel Machek 2005-01-06 23:41 ` Rafael J. Wysocki 2005-01-06 23:48 ` Pavel Machek 2005-01-07 0:24 ` Nigel Cunningham 2005-01-07 0:29 ` Pavel Machek 2005-01-07 0:45 ` Nigel Cunningham 2005-01-07 12:45 ` Rafael J. Wysocki 2005-01-07 22:12 ` Nigel Cunningham 2005-01-08 0:56 ` Rafael J. Wysocki 2005-01-08 9:49 ` 2.6.10-mm2: swsusp regression [update] Rafael J. Wysocki 2005-01-08 9:56 ` Nigel Cunningham 2005-01-08 13:19 ` Pavel Machek 2005-01-08 15:10 ` Rafael J. Wysocki 2005-01-08 15:44 ` Pavel Machek 2005-01-12 18:51 ` Rafael J. Wysocki 2005-01-12 21:01 ` Pavel Machek 2005-01-12 22:44 ` Rafael J. Wysocki 2005-01-12 22:46 ` Pavel Machek 2005-01-12 22:58 ` Nigel Cunningham 2005-01-12 23:02 ` Rafael J. Wysocki 2005-01-12 23:28 ` Nigel Cunningham 2005-01-13 0:59 ` [PATCH] Fix a bug in timer_suspend() on x86_64 Rafael J. Wysocki 2005-01-13 10:08 ` Pavel Machek 2005-01-13 10:14 ` Nigel Cunningham 2005-01-13 19:47 ` Andi Kleen 2005-01-08 17:22 ` 2.6.10-mm2: swsusp regression [update] Barry K. Nathan 2005-01-06 17:57 ` [patch] 2.6.10-mm2: move CPUSETS above EMBEDDED Adrian Bunk 2005-01-07 6:20 ` Paul Jackson 2005-01-06 18:15 ` [2.6 patch] 2.6.10-mm2: let I2C_ALGO_SGI depend on MIPS Adrian Bunk 2005-05-19 6:25 ` Adrian Bunk 2005-01-06 19:27 ` Ralf Baechle 2005-01-06 19:35 ` Ilya A. Volynets-Evenbakh 2005-05-19 6:25 ` Ilya A. Volynets-Evenbakh 2005-01-06 19:46 ` Greg KH 2005-05-19 6:25 ` Greg KH 2005-01-07 9:12 ` Ladislav Michl 2005-05-19 6:25 ` Ladislav Michl 2005-01-07 19:14 ` Greg KH 2005-01-06 20:30 ` 2.6.10-mm2 Ramón Rey Vicente 2005-01-07 9:39 ` 2.6.10-mm2 Benoit Boissinot 2005-01-07 13:13 ` 2.6.10-mm2 Brice Goglin 2005-01-09 1:26 ` 2.6.10-mm2 Dave Airlie 2005-01-09 8:55 ` 2.6.10-mm2 Brice Goglin 2005-01-08 2:43 ` 2.6.10-mm2 Dave Airlie 2005-01-08 12:27 ` 2.6.10-mm2 Benoit Boissinot 2005-01-08 13:42 ` 2.6.10-mm2 Dave Airlie 2005-01-08 14:18 ` 2.6.10-mm2 Marc Ballarin 2005-01-09 1:23 ` 2.6.10-mm2 Dave Airlie 2005-01-08 16:46 ` 2.6.10-mm2 Mike Werner 2005-01-08 13:48 ` 2.6.10-mm2 Dave Airlie 2005-01-08 15:41 ` 2.6.10-mm2 Benoit Boissinot 2005-01-09 1:38 ` 2.6.10-mm2 Dave Airlie 2005-01-09 14:09 ` 2.6.10-mm2 Benoit Boissinot 2005-01-09 14:15 ` 2.6.10-mm2 Brice Goglin 2005-01-10 7:40 ` 2.6.10-mm2 Dave Airlie 2005-01-10 7:49 ` 2.6.10-mm2 Brice Goglin [not found] ` <21d7e9970501100101353cf602@mail.gmail.com> 2005-01-10 10:14 ` 2.6.10-mm2 solved Brice Goglin 2005-01-10 16:41 ` Brice Goglin 2005-01-08 1:31 ` AGP Oops (was Re: 2.6.10-mm2) Sean Neakums 2005-01-08 1:36 ` Andrew Morton 2005-01-08 13:01 ` Sean Neakums 2005-01-06 14:06 ` [PATCH] macros to detect existance of unlocked_ioctl and ioctl_compat Michael S. Tsirkin 2005-01-06 14:53 ` Christoph Hellwig 2005-01-06 15:09 ` Andi Kleen 2005-01-06 15:14 ` Christoph Hellwig 2005-01-06 15:22 ` Lee Revell 2005-01-06 15:31 ` Christoph Hellwig 2005-01-06 15:38 ` Lee Revell [not found] ` <20050106154327.GA19781@infradead.org> 2005-01-06 16:21 ` new home for alsa-devel archive? (was Re: [PATCH] macros to detect existance of unlocked_ioctl and ioctl_compat) Lee Revell 2005-01-06 17:55 ` new home for alsa-devel archive? Måns Rullgård 2005-01-07 4:26 ` Eric Dantan Rzewnicki 2005-01-06 19:03 ` [PATCH] macros to detect existance of unlocked_ioctl and ioctl_compat Catalin Marinas 2005-01-06 15:37 ` Andi Kleen 2005-01-06 16:58 ` Petr Vandrovec 2005-01-06 16:35 ` Petr Vandrovec 2005-01-06 16:39 ` Christoph Hellwig 2005-01-06 16:57 ` Andi Kleen 2005-01-06 17:26 ` Petr Vandrovec 2005-01-06 17:53 ` Andi Kleen 2005-01-06 18:19 ` Petr Vandrovec 2005-01-06 19:35 ` Greg KH 2005-01-06 19:51 ` Andi Kleen 2005-01-06 19:59 ` David S. Miller 2005-01-06 20:44 ` Andi Kleen 2005-01-06 21:09 ` Petr Vandrovec 2005-01-06 21:24 ` Greg KH 2005-01-06 21:59 ` [linux-usb-devel] " David Brownell 2005-01-06 22:48 ` Greg KH 2005-01-07 11:49 ` [discuss] " Arnd Bergmann 2005-01-07 12:30 ` Andi Kleen 2005-01-06 22:34 ` Pavel Machek 2005-01-12 20:36 ` [PATCH] fix: macros to detect existance of unlocked_ioctl and compat_ioctl Michael S. Tsirkin 2005-01-12 21:29 ` Greg KH 2005-01-12 21:43 ` Andi Kleen 2005-01-12 22:52 ` Greg KH 2005-01-12 23:10 ` Andrew Morton 2005-01-12 23:16 ` Greg KH 2005-01-16 5:38 ` Werner Almesberger 2005-01-20 5:38 ` 2.6.11-rc1-mm2 Andrew Morton 2005-01-20 6:53 ` 2.6.11-rc1-mm2: CONFIG_SMP=n compile error Adrian Bunk 2005-01-20 7:12 ` Andrew Morton 2005-01-20 17:44 ` Rafael J. Wysocki 2005-01-20 9:57 ` 2.6.11-rc1-mm2 Steffen Klassert 2005-01-20 10:39 ` 2.6.11-rc1-mm2 Christoph Hellwig 2005-01-20 11:40 ` 2.6.11-rc1-mm2 Benoit Boissinot 2005-01-20 11:49 ` 2.6.11-rc1-mm2 Benoit Boissinot 2005-01-20 11:58 ` 2.6.11-rc1-mm2 Christoph Hellwig 2005-01-21 0:09 ` security hook missing in compat ioctl in 2.6.11-rc1-mm2 Michael S. Tsirkin 2005-01-21 0:10 ` Chris Wright 2005-01-21 1:26 ` [PATCH] compat ioctl security hook fixup Chris Wright 2005-01-21 4:19 ` Andi Kleen 2005-01-21 5:51 ` Chris Wright 2005-01-21 6:11 ` Andi Kleen 2005-01-21 6:35 ` [PATCH] compat ioctl security hook fixup (take2) Chris Wright 2005-01-21 8:06 ` 2.6.11-rc1-mm2 Con Kolivas 2005-01-21 8:30 ` 2.6.11-rc1-mm2 Andrew Morton 2005-01-21 8:34 ` 2.6.11-rc1-mm2 Andrew Morton 2005-01-21 8:37 ` 2.6.11-rc1-mm2 Con Kolivas 2005-01-21 12:48 ` 2.6.11-rc1-mm2 Adrian Bunk 2005-01-21 22:29 ` 2.6.11-rc1-mm2 Con Kolivas 2005-01-21 22:43 ` [PATCH] sched: account rt_tasks as iso_ticks Con Kolivas 2005-02-01 0:34 ` 2.6.11-rc1-mm2 Ed Tomlinson 2005-02-03 16:36 ` 2.6.11-rc1-mm2 Alexander Nyberg 2005-01-12 22:13 ` [PATCH] fix: macros to detect existance of unlocked_ioctl and compat_ioctl Andreas Dilger 2005-01-12 22:58 ` Greg KH 2005-01-13 10:57 ` Takashi Iwai
Reply instructions: You may reply publicly to this message via plain-text email using any one of the following methods: * Save the following mbox file, import it into your mail client, and reply-to-all from there: mbox Avoid top-posting and favor interleaved quoting: https://en.wikipedia.org/wiki/Posting_style#Interleaved_style * Reply using the --to, --cc, and --in-reply-to switches of git-send-email(1): git send-email \ --in-reply-to=m18y74rfqs.fsf@clusterfs.com \ --to=nikita@clusterfs.com \ --cc=akpm@osdl.org \ --cc=hch@infradead.org \ --cc=linux-kernel@vger.kernel.org \ --cc=linux-mm@kvack.org \ --cc=pmarques@grupopie.com \ /path/to/YOUR_REPLY https://kernel.org/pub/software/scm/git/docs/git-send-email.html * If your mail client supports setting the In-Reply-To header via mailto: links, try the mailto: linkBe sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes, see mirroring instructions on how to clone and mirror all data and code used by this external index.