Re: [Xen-devel] Assertion 'l1e_get_pfn(MAPCACHE_L1ENT(hashent->idx)) == hashent->mfn' failed at domain_page.c:203

[Konrad Rzeszutek Wilk <konrad.wilk@xxxxxxxxxx>]

On Tue, Dec 03, 2013 at 11:02:06AM +0000, Jan Beulich wrote:
> >>> On 02.12.13 at 21:33, Konrad Rzeszutek Wilk <konrad.wilk@xxxxxxxxxx> 
> >>> wrote:
> > This is with Xen 4.4 (pulled today) when I build a kernel in dom0 and
> > have two guests launching at the same time. This is what I get:
> > 
> > (XEN) Assertion 'l1e_get_pfn(MAPCACHE_L1ENT(hashent->idx)) == hashent->mfn' 
> > failed at domain_page.c:203
> > 
> > and it blows up.
> 
> Must be something specific to your guest config, extra patches you
> have on top, or something else no-one else has seen so far.

The top is
commit 26b35b9ace97f433fcf4c5dfbdfb573d1075255f
Author: Matthew Daley <mattd@xxxxxxxxxxx>
Date:   Sun Dec 1 23:15:03 2013 +1300

    libxl: don't leak pcidevs in libxl_pcidev_assignable

with this change:

 tools/libxl/libxlu_cfg_l.c | 28 +++++++---------------------
 tools/libxl/libxlu_cfg_l.h | 18 ++----------------
 2 files changed, 9 insertions(+), 37 deletions(-)

(Don't actually know why it keeps on doing that, but everytime
I build I see that file changed).
> 
> > (XEN) Assertion 'l1e_get_pfn(MAPCACHE_L1ENT(hashent->idx)) == hashent->mfn' 
> > failed at domain_page.c:203
> > (XEN) ----[ Xen-4.4-unstable  x86_64  debug=y  Not tainted ]----
> > (XEN) CPU:    5
> > (XEN) RIP:    e008:[<ffff82d080161cd6>] unmap_domain_page+0x1a7/0x1ff
> > (XEN) RFLAGS: 0000000000010083   CONTEXT: hypervisor
> > (XEN) rax: ffff8300ba9c1380   rbx: 0000000000000015   rcx: ffff83042cb59000
> > (XEN) rdx: ffff8300ba9c13c8   rsi: 00000000002410a0   rdi: 0000000000000015
> > (XEN) rbp: ffff83042cb77db8   rsp: ffff83042cb77d98   r8:  ffff814100200080
> > (XEN) r9:  0000000000244dbd   r10: 0000000000000282   r11: 0000000000000246
> > (XEN) r12: 0000000000000000   r13: 0000000000000001   r14: 0000000000000001
> > (XEN) r15: 0000000000000000   cr0: 0000000080050033   cr4: 00000000000406f0
> > (XEN) cr3: 0000000244b6b000   cr2: 00000036f5800280
> > (XEN) ds: 0000   es: 0000   fs: 0000   gs: 0000   ss: e010   cs: e008
> > (XEN) Xen stack trace from rsp=ffff83042cb77d98:
> > (XEN)    0000000264a4c025 0000000000000001 0000000000000000 0000000000000001
> > (XEN)    ffff83042cb77ef8 ffff82d08017c6b7 ffff8300ba9c1000 0000000100000000
> > (XEN)    80000002410a0061 0000000000000000 0000000000000000 ffff830400000000
> > (XEN)    ffff88013fe4b950 ffff83042cb77ea8 00000000002410a0 00000000002410a0
> > (XEN)    ffff8300ba9c1000 0000000000000000 ffff820040015000 ffff820040015000
> > (XEN)    00007ff000000003 0000000000000001 ffff82e004821400 ffff83042cb59000
> > (XEN)    ffff83042cb77eb8 ffff83042cb70000 ffff83042cb70000 0000000500000000
> > (XEN)    ffff83042cb59000 ffff8300ba9c1000 ffff83042cb59000 0000000500000001
> > (XEN)    aaaaaaaaaaaaaaaa 00000000aaaaaaaa 0000000000000000 00000000aaaaaaaa
> > (XEN)    00000002410a0000 0000000264a4c025 ffff83042cb77f08 ffff8300ba9c1000
> > (XEN)    ffff880116eac000 0000000264a4c025 0000000000000000 0000000264a4c025
> > (XEN)    00007cfbd34880c7 ffff82d0802248db ffffffff8100102a 0000000000000001
> > (XEN)    000000006ffffeff 0000000000000003 00007fffc2509190 00007f032448c4c8
> > (XEN)    ffff880069d13c98 0000000000000000 0000000000000246 0000000000007ff0
> > (XEN)    ffffea0002ea401c 0000000000000000 0000000000000001 ffffffff8100102a
> > (XEN)    0000000000000000 0000000000000001 ffff880069d13c60 0001010000000000
> > (XEN)    ffffffff8100102a 000000000000e033 0000000000000246 ffff880069d13c48
> > (XEN)    000000000000e02b ffffffffffffbeef ffffffffffffbeef ffffffffffffbeef
> > (XEN)    ffffffffffffbeef ffffffff00000005 ffff8300ba9c1000 00000033ac86b080
> > (XEN)    ffffffffffffffff
> > (XEN) Xen call trace:
> > (XEN)    [<ffff82d080161cd6>] unmap_domain_page+0x1a7/0x1ff
> > (XEN)    [<ffff82d08017c6b7>] do_mmu_update+0x16e3/0x19aa
> > (XEN)    [<ffff82d0802248db>] syscall_enter+0xeb/0x145
> 
> In any event, in order to make sense of this and the following one,
> I'd need to matching xen-syms.

Of course. Please see attached. If it didn't make it through I've
also put it on http://darnok.org/xen-syms.gz

> 
> Jan
>

Attachment: xen-syms.gz
Description: application/gzip

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
http://lists.xen.org/xen-devel