Subrahmanian, Raj
2005-Sep-22 20:43 UTC
[Xen-devel] Problems creating DomUs with large memory system/PAEenabled
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 the unstable tree : ) with PAE enabled (XEN_TARGET_X86_PAE=y). The system boots fine. I started xend and when I tried to bring up a virtual machine, the system crashed. This is what I see on the serial output. Raj (XEN) (file=traps.c, line=960) Non-priv domain attempted RDMSR(c0000080,00000000,00100000). (XEN) (file=traps.c, line=952) Non-priv domain attempted WRMSR(c0000080,00000800,00000000). (XEN) CPU: 3 (XEN) EIP: e008:[<ff1415b8>] get_page+0xd8/0x11a (XEN) EFLAGS: 00010282 CONTEXT: hypervisor (XEN) eax: 80000001 ebx: 80000002 ecx: ffbe2080 edx: ffbe2080 (XEN) esi: f7c57c7c edi: 00000000 ebp: ff1fddd8 esp: ff1fdd90 (XEN) cr0: 8005003b cr3: e742e000 (XEN) ds: e010 es: e010 fs: 0000 gs: 0000 ss: e010 cs: e008 (XEN) Xen stack trace from esp=ff1fdd90: (XEN) 00000000 ff11ec69 ffbe1080 00000038 00000067 00000001 fed6b020 ff1fddfc (XEN) ffbe2080 00000000 ffbe2080 ffbe2080 ffbe2080 80000001 80000002 80000001 (XEN) fde00000 0042e008 ff1fddf8 ff141630 f7c57c70 ffbe2080 ff1fde58 00000286 (XEN) 0000016b ff1fddfc ff1fde38 ff13945f f7c57c70 ffbe2080 e0000000 000003a9 (XEN) 00000004 000003aa 000003a9 ff18f56c 80000063 ffbe2088 f7c57c70 000d8fda (XEN) d8fda063 80000000 ff1fde78 ff13f609 d8fda063 80000000 ffbe2080 ff144cf7 (XEN) ff1fde80 00000020 ff1fde78 ff14483d 00000000 00000001 d8fda063 80000000 (XEN) 00000000 00000000 ff1fdec8 ff13f90b ffbe2080 fed6d000 ffbad000 000003a6 (XEN) e7494061 00000000 e7494063 00000000 00000000 00000000 ffbfa000 fde00000 (XEN) e7494061 00000000 fed6d000 c0000005 0042e008 fdfef000 ff1fdf48 ff14032a (XEN) ffbe2080 00000000 00000008 00000004 e742c067 00000000 000e742d 00000000 (XEN) 00000000 00000001 00000000 00000000 00000008 fde02170 ff1fdf28 00000600 (XEN) 00000001 00000000 e742c067 00000000 00000000 fdfef000 e742c021 00000000 (XEN) f7dae420 000e742c fed6a000 ffbe1080 0042e008 00000000 ff1fdf78 ff148222 (XEN) ffbe2080 c042e00c ff1fdfb4 00000001 00000000 f7dae438 00000001 00000003 (XEN) ffbe2080 ffbe1080 ff1fdfa8 ff1482a7 c042e00c ff1fdfb4 00000001 00e02037 (XEN) ff15cd6f c0355ea4 00000001 00000001 00007ff0 c042e00c 00e02037 ff15d002 (XEN) ff1fdfb4 ffffffff e7492063 00000000 0042e008 00000000 c0001000 e7492063 (XEN) 000e0003 c0115af6 0000e019 00010202 c0355ea0 0000e021 0000e021 0000e021 (XEN) 00000000 00000000 00000003 ffbe1080 (XEN) Xen call trace: (XEN) [<ff1415b8>] get_page+0xd8/0x11a (XEN) [<ff141630>] get_page_and_type+0x18/0x4e (XEN) [<ff13945f>] get_page_from_l1e+0x273/0x298 (XEN) [<ff13f609>] revalidate_l1+0x109/0x1a4 (XEN) [<ff13f90b>] ptwr_flush+0x267/0x32f (XEN) [<ff14032a>] ptwr_do_page_fault+0x2fc/0x57c (XEN) [<ff148222>] fixup_page_fault+0x1ef/0x256 (XEN) [<ff1482a7>] do_page_fault+0x1e/0x104 (XEN) [<ff15d002>] error_code+0x62/0x7d (XEN) (XEN) (XEN) **************************************** (XEN) Panic on CPU 3: (XEN) CPU3 FATAL TRAP: vector = 17 (alignment check) (XEN) [error_code=0000] (XEN) **************************************** (XEN) (XEN) Reboot in five seconds... _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel
Keir Fraser
2005-Sep-22 20:53 UTC
Re: [Xen-devel] Problems creating DomUs with large memory system/PAEenabled
On 22 Sep 2005, at 21:43, Subrahmanian, Raj wrote:> I started xend and when I tried to bring up a virtual machine, the > system crashed. > This is what I see on the serial output.Please dig deeper if you can. Something is odd here -- you cannot crash on an ''alignment check'' in Xen, as that fault can only occur in ring 3, and with the AC flag set in EFLAGS (which it isn''t). That''s got to be worth tracking down if this problem is reproducible for you. -- Keir _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel
David F Barrera
2005-Sep-22 21:09 UTC
Re: [Xen-devel] Problems creating DomUs with large memory system/PAEenabled
I have a similar system, IBM xSeries 335, 4 procs and 4GB memory, with SLES 9, PAE enabled, which is working fine. I am currently running test workloads on both dom0 and domU. changeset: 7029:43edce5b75d6 tag: tip user: kaf24@firebug.cl.cam.ac.uk date: Thu Sep 22 14:59:31 2005 +0100 summary: Remove file from sparse tree. Subrahmanian, Raj wrote:>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 the unstable tree : ) with PAE enabled >(XEN_TARGET_X86_PAE=y). >The system boots fine. > >I started xend and when I tried to bring up a virtual machine, the >system crashed. >This is what I see on the serial output. > >Raj > >(XEN) (file=traps.c, line=960) Non-priv domain attempted >RDMSR(c0000080,00000000,00100000). >(XEN) (file=traps.c, line=952) Non-priv domain attempted >WRMSR(c0000080,00000800,00000000). >(XEN) CPU: 3 >(XEN) EIP: e008:[<ff1415b8>] get_page+0xd8/0x11a >(XEN) EFLAGS: 00010282 CONTEXT: hypervisor >(XEN) eax: 80000001 ebx: 80000002 ecx: ffbe2080 edx: ffbe2080 >(XEN) esi: f7c57c7c edi: 00000000 ebp: ff1fddd8 esp: ff1fdd90 >(XEN) cr0: 8005003b cr3: e742e000 >(XEN) ds: e010 es: e010 fs: 0000 gs: 0000 ss: e010 cs: e008 >(XEN) Xen stack trace from esp=ff1fdd90: >(XEN) 00000000 ff11ec69 ffbe1080 00000038 00000067 00000001 fed6b020 >ff1fddfc >(XEN) ffbe2080 00000000 ffbe2080 ffbe2080 ffbe2080 80000001 80000002 >80000001 >(XEN) fde00000 0042e008 ff1fddf8 ff141630 f7c57c70 ffbe2080 ff1fde58 >00000286 >(XEN) 0000016b ff1fddfc ff1fde38 ff13945f f7c57c70 ffbe2080 e0000000 >000003a9 >(XEN) 00000004 000003aa 000003a9 ff18f56c 80000063 ffbe2088 f7c57c70 >000d8fda >(XEN) d8fda063 80000000 ff1fde78 ff13f609 d8fda063 80000000 ffbe2080 >ff144cf7 >(XEN) ff1fde80 00000020 ff1fde78 ff14483d 00000000 00000001 d8fda063 >80000000 >(XEN) 00000000 00000000 ff1fdec8 ff13f90b ffbe2080 fed6d000 ffbad000 >000003a6 >(XEN) e7494061 00000000 e7494063 00000000 00000000 00000000 ffbfa000 >fde00000 >(XEN) e7494061 00000000 fed6d000 c0000005 0042e008 fdfef000 ff1fdf48 >ff14032a >(XEN) ffbe2080 00000000 00000008 00000004 e742c067 00000000 000e742d >00000000 >(XEN) 00000000 00000001 00000000 00000000 00000008 fde02170 ff1fdf28 >00000600 >(XEN) 00000001 00000000 e742c067 00000000 00000000 fdfef000 e742c021 >00000000 >(XEN) f7dae420 000e742c fed6a000 ffbe1080 0042e008 00000000 ff1fdf78 >ff148222 >(XEN) ffbe2080 c042e00c ff1fdfb4 00000001 00000000 f7dae438 00000001 >00000003 >(XEN) ffbe2080 ffbe1080 ff1fdfa8 ff1482a7 c042e00c ff1fdfb4 00000001 >00e02037 >(XEN) ff15cd6f c0355ea4 00000001 00000001 00007ff0 c042e00c 00e02037 >ff15d002 >(XEN) ff1fdfb4 ffffffff e7492063 00000000 0042e008 00000000 c0001000 >e7492063 >(XEN) 000e0003 c0115af6 0000e019 00010202 c0355ea0 0000e021 0000e021 >0000e021 >(XEN) 00000000 00000000 00000003 ffbe1080 >(XEN) Xen call trace: >(XEN) [<ff1415b8>] get_page+0xd8/0x11a >(XEN) [<ff141630>] get_page_and_type+0x18/0x4e >(XEN) [<ff13945f>] get_page_from_l1e+0x273/0x298 >(XEN) [<ff13f609>] revalidate_l1+0x109/0x1a4 >(XEN) [<ff13f90b>] ptwr_flush+0x267/0x32f >(XEN) [<ff14032a>] ptwr_do_page_fault+0x2fc/0x57c >(XEN) [<ff148222>] fixup_page_fault+0x1ef/0x256 >(XEN) [<ff1482a7>] do_page_fault+0x1e/0x104 >(XEN) [<ff15d002>] error_code+0x62/0x7d >(XEN) >(XEN) >(XEN) **************************************** >(XEN) Panic on CPU 3: >(XEN) CPU3 FATAL TRAP: vector = 17 (alignment check) >(XEN) [error_code=0000] >(XEN) **************************************** >(XEN) >(XEN) Reboot in five seconds... > > > >_______________________________________________ >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
Puthiyaparambil, Aravindh
2005-Sep-23 18:39 UTC
RE: [Xen-devel] Problems creating DomUs with large memory system/PAEenabled
Keir, I am seeing the same issue with x86_64 Xen on the ES7000. We are looking into the problem. Aravindh (XEN) (file=traps.c, line=960) Non-priv domain attempted RDMSR(00000000c0000080,00020000,00020000). (XEN) CPU: 3 (XEN) RIP: e010:[<ffff830000123cff>] get_page_from_l1e+0x1af/0x4e0 (XEN) RFLAGS: 0000000000010246 CONTEXT: hypervisor (XEN) rax: 0000000080000002 rbx: 0000000080000003 rcx: 00000000001f7080 (XEN) rdx: 00000000001f7080 rsi: 0000000080000002 rdi: 000000000047e825 (XEN) rbp: 00000000001f7080 rsp: ffff8300001dfdb8 r8: 00000000001f7080 (XEN) r9: 00000000000e61a3 r10: 0fffffffe0000000 r11: ffff830000196680 (XEN) r12: 0000000080000002 r13: ffff8284023f4128 r14: ffff8300001f7080 (XEN) r15: ffff8300001f7080 cr0: 000000008005003b cr3: 00000000e7ea0000 (XEN) Xen stack trace from rsp=ffff8300001dfdb8: (XEN) ffff830000184ca0 ffff83000012000b ffff8300e61a3600 00000000e61a1065 (XEN) 00000000e61a1067 ffff8300001f7080 000000000047e82f ffff830000124d31 (XEN) ffff8284023f4178 00000000e61a1065 ffff8284023f4178 ffff8300e61a3600 (XEN) ffff828400000000 ffff8300001294e8 ffff8300001f4080 000000000007a120 (XEN) 0000000000000000 0000000000000002 ffff8284021b1f50 ffff8300001f7080 (XEN) ffff8300001f7080 00000000000e61a3 ffff8300001dff28 ffff8300001f4080 (XEN) 00000000ffc03ff8 0000000000000003 0000000000000000 00000000000e61a3 (XEN) 00007ff000000003 0000000000000000 00000001001f7080 ffffffff80393e40 (XEN) ffffffff807ff000 ffff8300001286c5 000000000000002c 0000000000183900 (XEN) 00000000e61a3600 00000000e61a1065 ffff8300001f4080 ffff8300001f4080 (XEN) 0000000000000600 ffffffff8034696d ffffffff802b0af3 ffffffff802b0ae3 (XEN) ffff880000000000 ffff83000013be3c ffff880000000000 ffffffff802b0ae3 (XEN) ffffffff802b0af3 ffffffff8034696d 0000000000000600 00000000e61a3000 (XEN) 0000000000000206 0000000000007ff0 0000000000000000 0000000000000000 (XEN) 0000000000000001 ffffffff8011caa1 0000000000000000 0000000000000001 (XEN) ffffffff80393e40 0000010000000000 ffffffff8011caa1 000000000000e033 (XEN) 0000000000000206 ffffffff80393e40 000000000000e02b 0000000000000000 (XEN) 0000000000000000 0000000000000000 0000000000000000 0000000000000003 (XEN) ffff8300001f4080 (XEN) Xen call trace: (XEN) [<ffff830000123cff>] get_page_from_l1e+0x1af/0x4e0 (XEN) [<ffff83000012000b>] apic_timer_interrupt+0x2b/0x30 (XEN) [<ffff830000124d31>] mod_l1_entry+0x121/0x2d0 (XEN) [<ffff8300001294e8>] do_mmu_update+0x808/0x1a00 (XEN) [<ffff8300001286c5>] do_update_va_mapping+0xc5/0x5f0 (XEN) [<ffff83000013be3c>] syscall_enter+0x5c/0x61 (XEN) (XEN) (XEN) **************************************** (XEN) Panic on CPU 3: (XEN) CPU3 FATAL TRAP: vector = 17 (alignment check) (XEN) [error_code=0000] (XEN) ****************************************> -----Original Message----- > From: Keir Fraser [mailto:Keir.Fraser@cl.cam.ac.uk] > Sent: Thursday, September 22, 2005 4:54 PM > To: Subrahmanian, Raj > Cc: Vessey, Bruce A; Puthiyaparambil, Aravindh; xen- > devel@lists.xensource.com; Koren, Bradley J > Subject: Re: [Xen-devel] Problems creating DomUs with large memory > system/PAEenabled > > > On 22 Sep 2005, at 21:43, Subrahmanian, Raj wrote: > > > I started xend and when I tried to bring up a virtual machine, the > > system crashed. > > This is what I see on the serial output. > > Please dig deeper if you can. Something is odd here -- you cannotcrash> on an ''alignment check'' in Xen, as that fault can only occur in ring3,> and with the AC flag set in EFLAGS (which it isn''t). > > That''s got to be worth tracking down if this problem is reproducible > for you. > > -- Keir_______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel