similar to: RE: [PATCH] supporting non-NX/XD capable processors onx86_64 xenlinux

The patched attached enables x86_64 xenlinux with "late pin, early unpin", which is already implemented for x86_32. Since we now only pin the root rather than any of the other levels, the overall performance became better especially with workloads that require heavy memory management operations. On 8-way x86_64 xenlinux (dom0) the kernel build was improved by about 10% (using make

Whether the processor is in 32 or 64-bit mode, if NX is used, then EFER_NX needs to be set. If NX isn''t used, then it''s a "Don''t care". I think bad things happens if you set the NX bit in the page table and don''t have EFER_NX set... -- Mats > -----Original Message----- > From: xen-devel-bounces@lists.xensource.com >

Xen patchbot -unstable <patchbot-unstable@lists.xensource.com> writes: > Fix NX/XD enable on secondary CPUs. > Signed-off-by: Keir Fraser <keir@xensource.com> I think I have this problem with PAE as well. Machine is SMP (hyperthreaded). PAE dom0 boots fine on CPU #0. PAE domU is bound to CPU #1 by default and boots to the login prompt as well, but only with NX disabled (and

Yes, like real Linux/ia64, it is not possible to run the same Xenlinux bits both on ski and on real hardware. A different configuration and build is required. The config difference is the same as it was with the privified xenlinux... I think you did that successfully before (with privified xenlinux) so give that config file a try and if it doesn''t work, let me know. BTW, I think I know

BTW, the xenlinux-2.6.11.bk''s version is: ChangeSet@1.2085, 2005-05-12 11:10:24-06:00, djm@kirby.fc.hp.com More work on hyperprivops Signed-off-by: Dan Magenheimer <dan.magenheimer@hp.com> Thanks, Kevin >-----Original Message----- >From: xen-devel-bounces@lists.xensource.com >[mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Tian, Kevin >Sent: Friday,

Hi. I forward ported XenLinux/IA64 domU portion and put it at http://people.valinux.co.jp/~yamahata/xen-ia64/20080207/xen_ia64-20080207.patch This patch is for 551e4fb2465b87de9d4aa1669b27d624435443bb of git://git.kernel.org/pub/scm/linux/kernel/git/aegl/linux-2.6.git It boots as domU and disk/network work. This patch is for those who are interested in XenLinux/IA64 upstream merge activity. Many

Hi, Keir, Any comments upon following patch? ;-) We need it to further check in a specific configuration file (likely xen0_defconfig_ia64tiger) later to run xen on IA64 tiger box. Currently we have no generic one for all IA64 boxes. Let''s know if better way to do it. Currently custom with tiger box can only get that file out of the tree. Thanks, Kevin >-----Original Message-----

>From: Keir Fraser >Sent: 2006年4月28日 21:57 >On 28 Apr 2006, at 10:14, Isaku Yamahata wrote: > >> Unfortunately no. >> Roughly init_mm is only used to map vmalloc area on Linux/ia64. >> init_mm of Linux/ia64 doesn''t map the area of [PAGE_OFFSET, ...]. >> Traversing init_mm with a virtual address of the area >> gives a zero-filled pte entry.

On Fri, May 14, 2010 at 09:10:57PM +0200, Peter Klar wrote: > As the bug seems to be related to the SLAB allocator, the dump says ''kernel > BUG at mm/slub.c:2969!'', I also recompiled the kernel using the SLAB instead > of SLUB allocator, but this does not make any difference, the behaviour is > the same (beside the dump then reports a bug within slab.c instead of

On Fri, May 14, 2010 at 09:10:57PM +0200, Peter Klar wrote: > As the bug seems to be related to the SLAB allocator, the dump says ''kernel > BUG at mm/slub.c:2969!'', I also recompiled the kernel using the SLAB instead > of SLUB allocator, but this does not make any difference, the behaviour is > the same (beside the dump then reports a bug within slab.c instead of

Hi list! I''m trying to run 2.4.29-xenU with grsec. Jacob Gorm Hansen said couple of weeks ago, that grsec should work with xen when pax is disabled.. Well, to get the kernel compiling there''s some source hacking that needs to be done.. I''ll describe what I did and what error I got: I downloaded xen-2.0-testing-src.tgz and extracted it. I edited the toplevel Makefile

Hello, I''m trying to build XenLinux 2.4.30 on debian. I get the following error for make linux24 make[2]: *** No rule to make target `skbuff.c'',needed by `/usr/src/xen-2.0/linux-2.4.30-xen0/include/linux/modules/skbuff.ver''. Stop. Could somebody help me resolve this error. Or if this is some problem with the system configuration that I''m using, could

The DMI checks that force the use of the bigsmp APIC driver are for systems that are no longer supported by Xen (32-bit x86). Signed-off-by: Matt Wilson <msw@amazon.com> Cc: Keir Fraser <keir@xen.org> Cc: Jan Beulich <jbeulich@suse.com> Cc: Andrew Cooper <andrew.cooper3@citrix.com> --- xen/arch/x86/genapic/bigsmp.c | 30 +----------------------------- 1 files changed, 1

On Unisys 16X 16GB ES7000/one with SLES10 Beta11 updated to xen changeset 9998, attempted to create a 2GB domU and get Error: The privileged domain did not balloon. xm list shows that before the attempt to create the domain, dom0 has 15910 MB memory. After the error, dom0 has dropped down to 13993 MB. Bugzilla number is 650 > Sue Krysan > Linux Systems Group > Unisys Corporation >

Hi, The setup used to work on a when both file and AD were in the same box. We're trying to separate them. The 'net rpc testjoin' gives: Join to "LIN" is ok. The wbinfo -u does list all users with "LIN\username". The getent passwd lists the LIN\username with all the attrributes". This is after putting in your suggestions about winbind trusted domains only and

Here is a patch for clustered xAPIC and Unisys ES7000 support in Xen. To turn the features on add CONFIG_X86_CLUSTERED_APIC and CONFIG_ES7000 defines to xen/include/asm-x86/config.h Signed off by Aravindh Puthiyaparambil <aravindh.puthiyaparambil@unisys.com> _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com

With this patch, x86 and x86-64 share ioport.c, fixing LTP iopl and ioperm testcase failures (on both). We found an iopl testcase failing even on x86 xenlinux. Now x86-64 xenlinux should have the same results regarding the LTP testcases (as far as we tested). Signed-off-by: Li B Xin <li.b.xin@intel.com> Signed-off-by: Jun Nakajima <jun.nakajima@intel.com> Jun --- Intel Open Source

Hi Raj, I also meet the same python problem when create VMX in FC3-x86_64 xen0. But python error doesn''t happen in RHEL4U1-x86_64 xen0. Are your two computers all SLES9? Thanks, Yongkang You(Kangkang) > >All. > >I am running Xen on an ES7000 system with 4 procs and 4 GB memory. >I am running Suse Sles-9 (32-bit) on the machine. > >I have compiled Xen (the tip on

Using 3.1.2 c/s 15492 on Unisys ES7000 configured with 32 processors and 128 gb memory, I attempted to start a 160 gb domU. The xm create command hangs, as well as any other xm commands that are issued. If I do ctl-c of the hanging xm create command, I am able to terminate it and then xm commands are once again functional. I expected to get an error that there was not enough memory available

The host server halts upon shutdown of 126000mb domU (64-bit SLES10). It booted fine and was able to run kernbench. However, during shutdown the host crashes. Host is Unisys ES7000 host with 32 processors and 128gb memory using x86_64 SLES10 with xen-unstable c/s 15080. Reducing the memory allocation for the domU to 124000mb works fine. My guess is that specifying 126000 does not leave