On Sat, January 6, 2018 3:20 pm, awokd wrote: > Trying to debug an HVM PCI pass-through crash coming from: > > > https://xenbits.xen.org/gitweb/?p=xen.git;a=blob;f=xen/arch/x86/hvm/svm/s > vm.c;h=8edc846a2ed4aec6a5711d272adbb10aaa741c27;hb=5e4598106ed02ae4b43abc > b29889969eb12867b7 1536 /* Everything else is an error. */ > 1537 mfn = __get_gfn_type_access(p2m, gfn, &p2mt, &p2ma, 0, NULL, 0); > 1538 gdprintk(XENLOG_ERR, > 1539 "SVM violation gpa %#"PRIpaddr", mfn %#lx, type %i\n", > 1540 gpa, mfn_x(mfn), p2mt); > 1541 domain_crash(v->domain); > > > Which results in: > > > (XEN) domain_crash called from svm.c:1541 > (XEN) Domain 10 (vcpu#0) crashed on cpu#2: > (XEN) ------[ Xen-4.8.2 x86_64 debug=n Not tainted ]------ > (XEN) CPU: 2 > (XEN) RIP: 0010:[<ffffffff97405f4e>] > (XEN) RFLAGS: 0000000000000296 CONTEXT: hvm guest (d10v0) > (XEN) rax: 0000000000000000 rbx: ffff97c946eb95c0 rcx: > 0000000000000005 > (XEN) rdx: 0000000000000040 rsi: ffffaf5580700040 rdi: > 0000000000000000 > (XEN) rbp: ffffaf55806cb8f8 rsp: ffffaf55806cb8c8 r8: > 0000000000000000 > (XEN) r9: 00000000ffffff90 r10: 000000000000003f r11: > 0000000000000000 > (XEN) r12: 0000000000000000 r13: ffffffffc04fb7d0 r14: > 0000000000000100 > (XEN) r15: ffff97c946eb4028 cr0: 0000000080050033 cr4: > 00000000000406f0 > (XEN) cr3: 0000000012e30000 cr2: 0000000000000000 > (XEN) ds: 0000 es: 0000 fs: 0000 gs: 0000 ss: 0018 cs: 0010 > > > Here's where I get lost. Can't find the output from that gdprintk > anywhere, with both loglvl=all and guest_loglvl=all. I think I want to > either change it to a printk and/or enable Xen debugging, which means a > recompile. What is the most newbish/straight-forward way to do that? > Looks > like Qubes-builder compiles Xen in > chroot-fc25/home/user/rpmbuild/BUILD/xen-4.8.2. I can find and edit > svm.c, but can't figure out where I should enable Xen debugging and what > to do next.
P.S. Posting this here instead of one of the Xen lists because this pass-through works without crashing on a stock Fedora 26/Xen 4.8.2 boot on the same hardware. There are more details in my Qubes-users thread (https://mail-archive.com/[email protected]/msg17452.html) but I'm hoping I can get this narrowed down a bit more myself! -- You received this message because you are subscribed to the Google Groups "qubes-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To post to this group, send email to [email protected]. To view this discussion on the web visit https://groups.google.com/d/msgid/qubes-devel/8dacfda2e616224f4ebf0809c3e85a12.squirrel%40tt3j2x4k5ycaa5zt.onion. For more options, visit https://groups.google.com/d/optout.
