Re: [patch 1/3] powerpc: Move 64bit VDSO to improve context switch performance

2009-10-04 Thread Andreas Schwab
Benjamin Herrenschmidt writes: > Maybe a better fix is to force alignment in the kernel by requesting > size + 64k - 4k and aligning it. Sure you are right. Andreas. >From b9441a3d2148d439e2730def3222a7b70dccc432 Mon Sep 17 00:00:00 2001 From: Andreas Schwab Date: Sun, 4 Oct 2009 14:29:04 +02

Re: [patch 1/3] powerpc: Move 64bit VDSO to improve context switch performance

2009-10-03 Thread Benjamin Herrenschmidt
On Sat, 2009-10-03 at 16:51 +0200, Andreas Schwab wrote: > Andreas Schwab writes: > > > Andreas Schwab writes: > > > >> Anton Blanchard writes: > >> > >>> On 64bit applications the VDSO is the only thing in segment 0. Since the > >>> VDSO > >>> is position independent we can remove the hint an

Re: [patch 1/3] powerpc: Move 64bit VDSO to improve context switch performance

2009-10-03 Thread Andreas Schwab
Andreas Schwab writes: > Andreas Schwab writes: > >> Anton Blanchard writes: >> >>> On 64bit applications the VDSO is the only thing in segment 0. Since the >>> VDSO >>> is position independent we can remove the hint and let get_unmapped_area >>> pick >>> an area. >> >> This breaks gdb. The

Re: [patch 1/3] powerpc: Move 64bit VDSO to improve context switch performance

2009-10-03 Thread Andreas Schwab
Andreas Schwab writes: > Anton Blanchard writes: > >> On 64bit applications the VDSO is the only thing in segment 0. Since the VDSO >> is position independent we can remove the hint and let get_unmapped_area pick >> an area. > > This breaks gdb. The section table in the VDSO image when mapped i

Re: [patch 1/3] powerpc: Move 64bit VDSO to improve context switch performance

2009-10-02 Thread Andreas Schwab
Anton Blanchard writes: > On 64bit applications the VDSO is the only thing in segment 0. Since the VDSO > is position independent we can remove the hint and let get_unmapped_area pick > an area. This breaks gdb. The section table in the VDSO image when mapped into the process no longer contains

Re: [patch 1/3] powerpc: Move 64bit VDSO to improve context switch performance

2009-07-14 Thread Anton Blanchard
Hi Ben, > Don't we lose randomization ? Or do we randomize the whole mem map > nowadays ? The start of the top down mmap region is randomized, so the VDSO will be in a different position each time. A quick example: run 1: fffb01f6000-fffb01f9000 r-xp 00:00 0 [vdso] fffb01f9000

[patch 1/3] powerpc: Move 64bit VDSO to improve context switch performance

2009-07-14 Thread Anton Blanchard
On 64bit applications the VDSO is the only thing in segment 0. Since the VDSO is position independent we can remove the hint and let get_unmapped_area pick an area. This will mean the vdso will be near other mmaps and will share an SLB entry: 1000-10001000 r-xp 08:06 5778459/r

Re: [patch 1/3] powerpc: Move 64bit VDSO to improve context switch performance

2009-07-14 Thread Benjamin Herrenschmidt
On Tue, 2009-07-14 at 16:53 +1000, Anton Blanchard wrote: > plain text document attachment (move_vdso_v2) > On 64bit applications the VDSO is the only thing in segment 0. Since the VDSO > is position independent we can remove the hint and let get_unmapped_area pick > an area. This will mean the vds