Bugs item #1842160, was opened at 2007-11-30 23:12 Message generated for change (Comment added) made by iggy_cav You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=893831&aid=1842160&group_id=180599 Please note that this message will contain a full copy of the comment thread, including the initial issue submission, for this request, not just the latest update. Category: intel Group: None >Status: Pending Resolution: Fixed Priority: 5 Private: No Submitted By: Carlo Marcelo Arenas Belon (carenas) Assigned to: Amit Shah (amitshah) Summary: Solaris 10 8/07 (AKA update 4) x86 guest can't boot 64 bit Initial Comment: Host CPU: Intel(R) Core(TM)2 CPU 6320 KVM: 54 + ide patches and set_cr3 patch pre-kvm-55 Host OS: Gentoo 2007.0 with kernel 2.6.22-gentoo-r9 Host ARCH: amd64 (64bit) Guest OS: Solaris 10 8/07 x86 Generic 120012-14 VM parameters: -m 512 problem only happens when running kvm's CPU emulation (doesn't happen with -no-kvm) and triggers a GP as shown by : set_cr4: #GP, clearing PAE while in long mode inject_general_protection: rip 0x100ee24. it might be related to the GP reported as part of the discussion for bug 1699695 which apparently has a patch available (but not yet committed) which I hadn't been able to track down to test ---------------------------------------------------------------------- >Comment By: Brian Jackson (iggy_cav) Date: 2010-02-08 14:33 Message: Does still occur with the latest kvm kernel module and userspace? ---------------------------------------------------------------------- Comment By: Carlo Marcelo Arenas Belon (carenas) Date: 2008-07-13 01:18 Message: Logged In: YES user_id=36771 Originator: YES This still affects the last update (u5) of Solaris 10 (5/08) but interestingly enough does not affect OpenSolaris versions based in nevada b85 (OpenSolaris 2008.05 or Nexenta Core 2.0 alpha 1) ---------------------------------------------------------------------- Comment By: Amit Shah (amitshah) Date: 2008-03-04 06:08 Message: Logged In: YES user_id=201894 Originator: NO Looks like we'll have to wait more. windows 64-bit doesn't like this change. ---------------------------------------------------------------------- Comment By: Amit Shah (amitshah) Date: 2008-03-04 00:13 Message: Logged In: YES user_id=201894 Originator: NO Fix is in kvm-git. Will be part of kvm-63. ---------------------------------------------------------------------- Comment By: Carlo Marcelo Arenas Belon (carenas) Date: 2007-12-01 03:02 Message: Logged In: YES user_id=36771 Originator: YES there was a bug in the way I disabled the GP for set_cr4 which triggered the set_cr0, so that is not a real issue. the suggested patch fixes the problem (using an Intel Host as well as it does most likely for AMD hosts as reported in the thread). will leave bug open for now, to help for follow up into committing this patch and release ---------------------------------------------------------------------- Comment By: Amit Shah (amitshah) Date: 2007-12-01 02:12 Message: Logged In: YES user_id=201894 Originator: NO Use the patch in http://www.mail-archive.com/kvm-devel%40lists.sourceforge.net/msg08947.html to resolve the #gp ---------------------------------------------------------------------- Comment By: Carlo Marcelo Arenas Belon (carenas) Date: 2007-12-01 01:20 Message: Logged In: YES user_id=36771 Originator: YES if LME/LMA handling is considered buggy as reported in BUG 1699695 and the GP commented out, another GP is generated after as shown by : BUG: set_cr4: #GP, clearing PAE while in long mode set_cr0: #GP, pdptrs reserved bits inject_general_protection: rip 0x100ee49 ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=893831&aid=1842160&group_id=180599 -- To unsubscribe from this list: send the line "unsubscribe kvm" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
