Puthiyaparambil, Aravindh
2005-May-18 20:44 UTC
RE: [Xen-devel] x86_64/entry.S:317: Error: `1(%eax)'' is not a valid64 bit base/index
So has the "prefetchw" bug been fixed? http://bugzilla.xensource.com/cgi-bin/bugzilla/show_bug.cgi?id=34. I am still seeing it on SLES9 SP1 x86_64. I am using the latest unstable pull. Aravindh -----Original Message----- From: xen-devel-bounces@lists.xensource.com [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Jerone Young Sent: Wednesday, May 18, 2005 3:57 PM To: Vincent Hanquez Cc: David F Barrera; xen-devel@lists.xensource.com Subject: Re: [Xen-devel] x86_64/entry.S:317: Error: `1(%eax)'' is not a valid64 bit base/index The fix for this is in bitkeeper now. It did not make it into the unstable tar ball for the night. So you will not see this problem tommorow. On 5/18/05, Vincent Hanquez <vincent.hanquez@cl.cam.ac.uk> wrote:> On Wed, May 18, 2005 at 10:59:00AM -0500, David F Barrera wrote: > > Using the May 17 xen-unstable-src.tgz on x86_64, the following error > > occurs on a SLES 9 SP1 platform (It builds OK on Fedora Core 4, butit> > does not boot, and that is another issue): > > > > gcc -nostdinc -fno-builtin -fno-common -fno-strict-aliasing-iwithprefix include> > -Wall -Werror -Wno-pointer-arith -pipe-I/tmp/xen-unstable/xen/include> > -I/tmp/xen-unstable/xen/include/asm-x86/mach-default -O3-fomit-frame-pointer> > -msoft-float -m64 -mno-red-zone -fpic -fno-reorder-blocks > > -fno-asynchronous-unwind-tables -DNDEBUG -D__ASSEMBLY__ -cx86_64/entry.S -o> > x86_64/entry.o > > x86_64/entry.S: Assembler messages: > > x86_64/entry.S:317: Error: `1(%eax)'' is not a valid 64 bitbase/index> > expressionx86_64/entry.S:320: Error: `136+8(%esp)'' is not a valid 64bit> > base/index expression > > that''s odd. could you verify you have the correct line on entry.S ? > > line 317: setnz VCPUINFO_upcall_mask(%rax) > line 320: movw UREGS_cs+8(%rsp),%ax > > as well, what compiler version are you using ? > > -- > Vincent Hanquez > > _______________________________________________ > Xen-devel mailing list > Xen-devel@lists.xensource.com > http://lists.xensource.com/xen-devel >_______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel
Jerone Young
2005-May-19 02:28 UTC
Re: [Xen-devel] x86_64/entry.S:317: Error: `1(%eax)'' is not a valid64 bit base/index
I looked into this problem breifely yesterday. This looks to be something with SuSEs gcc setup. Right now this is the least of our worries in x86-64 land (not that I''m not going to look hard into it and solve it). There are a lot of other problems with x86-64 Xen eating up my time right now. I worked on SLES9 bring up when it was coming to release. I can tell you they put a lot of crazy patches in their gcc that cause stuff like this. It''s a matter of figuring out what is triggering it. Because x86-64 Xen builds fine on Debian, Ubuntoo, Gentoo, & Fedora. On 5/18/05, Puthiyaparambil, Aravindh <aravindh.puthiyaparambil@unisys.com> wrote:> So has the "prefetchw" bug been fixed? > http://bugzilla.xensource.com/cgi-bin/bugzilla/show_bug.cgi?id=34. > > I am still seeing it on SLES9 SP1 x86_64. I am using the latest unstable > pull. > > Aravindh > > -----Original Message----- > From: xen-devel-bounces@lists.xensource.com > [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Jerone Young > Sent: Wednesday, May 18, 2005 3:57 PM > To: Vincent Hanquez > Cc: David F Barrera; xen-devel@lists.xensource.com > Subject: Re: [Xen-devel] x86_64/entry.S:317: Error: `1(%eax)'' is not a > valid64 bit base/index > > The fix for this is in bitkeeper now. It did not make it into the > unstable tar ball for the night. So you will not see this problem > tommorow. > > On 5/18/05, Vincent Hanquez <vincent.hanquez@cl.cam.ac.uk> wrote: > > On Wed, May 18, 2005 at 10:59:00AM -0500, David F Barrera wrote: > > > Using the May 17 xen-unstable-src.tgz on x86_64, the following error > > > occurs on a SLES 9 SP1 platform (It builds OK on Fedora Core 4, but > it > > > does not boot, and that is another issue): > > > > > > gcc -nostdinc -fno-builtin -fno-common -fno-strict-aliasing > -iwithprefix include > > > -Wall -Werror -Wno-pointer-arith -pipe > -I/tmp/xen-unstable/xen/include > > > -I/tmp/xen-unstable/xen/include/asm-x86/mach-default -O3 > -fomit-frame-pointer > > > -msoft-float -m64 -mno-red-zone -fpic -fno-reorder-blocks > > > -fno-asynchronous-unwind-tables -DNDEBUG -D__ASSEMBLY__ -c > x86_64/entry.S -o > > > x86_64/entry.o > > > x86_64/entry.S: Assembler messages: > > > x86_64/entry.S:317: Error: `1(%eax)'' is not a valid 64 bit > base/index > > > expressionx86_64/entry.S:320: Error: `136+8(%esp)'' is not a valid 64 > bit > > > base/index expression > > > > that''s odd. could you verify you have the correct line on entry.S ? > > > > line 317: setnz VCPUINFO_upcall_mask(%rax) > > line 320: movw UREGS_cs+8(%rsp),%ax > > > > as well, what compiler version are you using ? > > > > -- > > Vincent Hanquez > > > > _______________________________________________ > > Xen-devel mailing list > > Xen-devel@lists.xensource.com > > http://lists.xensource.com/xen-devel > > > > _______________________________________________ > Xen-devel mailing list > Xen-devel@lists.xensource.com > http://lists.xensource.com/xen-devel >_______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel