Hello!
Yesterday I tried to run the tests from tools/xm-test. This is the
first time that I spend some time checking these things - so some of
the questions and remarks might be just beginner ones (please be
forbearing). The background is, that I want to add a new feature -
and want at least somewhat sure that this breaks no existing
functionality. The following tests were run on plain xen-unstable.
(Most of the tests passed: 96 PASS / 13 FAIL / 21 SKIP)
Issues:
1) The initrd [1] comes with an ext2 file system. The pvops one used
in xen-unstable (2.6.30-rc3-tip) has by default ext2 disabled.
[My proposal: provide an initrd with ext3 fs.]
2) This kernel config also lacks bridge capabilities - also those
must be enabled to get xend with bridged network running.
[Is it possible to add this capability to the kernel config or
will this break other things?]
3) 02_block_attach_file_device_pos: failed with
[dom0] Running `xm block-attach \
02_block_attach_file_device_pos-1248776182 file:/dev/ram1
xvda1 w''
Error: Device 51713 (vbd) could not be connected. /dev/ram1 does
not exist.
Usage: xm block-attach <Domain> <BackDev> <FrontDev>
<Mode>
[BackDomain]
When trying to reproduce the failure to get some more information,
the guest system crashes (see Attachment 1 for a stack trace).
4) 05_block_attach_and_dettach_device_repeatedly_pos: one
loop works (one attach and detach). The second attach stops with:
Error: Device xvda1 (51713, vbd) is already connected.
5) 09_block_attach_and_dettach_device_check_data_pos:
see 4)
6) 10_block_attach_dettach_multiple_devices: is
available but disabled. [Why?]
7) 16_create_smallmem_neg: failed with
REASON: Unable to start a domain with 16 MB
8) 01_enforce_dom0_cpus_basic_pos:
failed with ''REASON: failed to restore dom0''s
VCPUs''
9) 01_memset_basic_pos and 03_memset_random_pos:
they are using /proc/xen/balloon which is not available in the
pvops kernel.
[A patch for this looks straigt forward: when available use the
sysfs.]
10) 04_memset_smallmem_pos: failed with
REASON: Timed out waiting for console prompt
11) 13_network_domU_udp_pos:
REASON: UDP hping2 failed for size 1 48 64 512 1440 \
1500 1505 4096 4192 32767 65495.
12) 02_network_attach_detach_pos:
REASON: Network device is not actually connected to domU
13) 03_network_attach_detach_multiple_pos:
REASON: Network device was not actually disconnected from domU
14) 04_restore_withdevices_pos:
REASON: Unable to mount /dev/hda1
15) When enabling xm Xen-API as described in the xm-test/README there
are problems starting guest systems. (see Attachment 2)
16) In the README there is written that reports are send to
''http://xmtest.dague.org'' but the domain name has no IP
address. (NXDOMAIN)
Can anybody please give me some hints? Are these known problems? Are
other people running in similar problems executing xm-tests?
Kind regards
Andreas Florath
[1] http://xm-test.xensource.com/ramdisks/initrd-1.1-i386.img
======================================== Attachment 1 =========[ 79.184978]
BUG: unable to handle kernel NULL pointer dereference at
0000002c
[ 79.185040] IP: [<c118a149>] disk_get_part+0x8/0x2d
[ 79.185089] *pdpt = 0000000005980001
[ 79.185124] Oops: 0000 [#1] SMP
[ 79.185165] last sysfs file:
[ 79.185186] Modules linked in:
[ 79.185217]
[ 79.185239] Pid: 15, comm: xenwatch Not tainted (2.6.30-rc3-tip #4)
[ 79.185267] EIP: 0061:[<c118a149>] EFLAGS: 00010246 CPU: 0
[ 79.185295] EIP is at disk_get_part+0x8/0x2d
[ 79.185320] EAX: 00000000 EBX: c30c5100 ECX: c30d8000 EDX: 00000000
[ 79.185348] ESI: 00000000 EDI: c1688818 EBP: c384df34 ESP: c384df30
[ 79.185375] DS: 007b ES: 007b FS: 00d8 GS: 0000 SS: 0069
[ 79.185401] Process xenwatch (pid: 15, ti=c384c000 task=c3848000
task.ti=c384c000)
[ 79.185430] Stack:
[ 79.185448] c30c5100 c384df44 c118a20f c30c5100 c30c5100 c384df80
c1222338 c15ff85a
[ 79.185542] c38ba000 c3064280 c1688818 c384df80 c11e1bf6 00000000
c3064100 c15e4265
[ 79.185648] c15ff85a c30c5100 c3064280 c1688818 c384df94 c11e4046
c30c5110 c3064380
[ 79.185767] Call Trace:
[ 79.185787] [<c118a20f>] ? bdget_disk+0xc/0x2d
[ 79.185817] [<c1222338>] ? backend_changed+0x38e/0x428
[ 79.185817] [<c11e1bf6>] ? xenbus_read_driver_state+0x1e/0x31
[ 79.185817] [<c11e4046>] ? xenbus_otherend_changed+0x5c/0x61
[ 79.185817] [<c11e44f9>] ? backend_changed+0xa/0xd
[ 79.185817] [<c11e2c6c>] ? xenwatch_thread+0xf9/0x121
[ 79.185817] [<c1061c57>] ? autoremove_wake_function+0x0/0x33
[ 79.185817] [<c11e2b73>] ? xenwatch_thread+0x0/0x121
[ 79.185817] [<c106183c>] ? kthread+0x45/0x6b
[ 79.185817] [<c10617f7>] ? kthread+0x0/0x6b
[ 79.185817] [<c1026717>] ? kernel_thread_helper+0x7/0x10
[ 79.185817] Code: 00 00 eb 06 89 b0 c0 00 00 00 89 d8 e8 8e fb ff ff
85 c0 75 ef 89 d8 e8 7e f6 ff ff 83 c4 10 5b 5e 5d c3 55 85 d2 89 e5 53
78 20 <8b> 40 2c 3b 50 08 7d 18 8d 44 90 10 8b 18 31 c0 85 db 74 0e 8d
[ 79.185817] EIP: [<c118a149>] disk_get_part+0x8/0x2d SS:ESP
0069:c384df30
[ 79.185817] CR2: 000000000000002c
[ 79.185817] ---[ end trace 3aeec7b7d74bc5c7 ]---
[ 79.187070] xenwatch used greatest stack depth: 6432 bytes left
======================================== Attachment 2 =========# xm create -c
testdomu2.sxp
Using config file "./testdomu2.sxp".
ERROR: Couldn''t open resource
''http://www.w3.org/TR/xhtml1/DTD/xhtml-lat1.ent'' at
/usr/share/create.dtd:4:11
ERROR: Couldn''t open resource
''http://www.w3.org/TR/xhtml1/DTD/xhtml-symbol.ent'' at
/usr/share/create.dtd:8:13
Could not start console
# ERROR: Couldn''t open resource
''http://www.w3.org/TR/xhtml1/DTD/xhtml-special.ent'' at
/usr/share/create.dtd:12:14
xm list
Name ID Mem VCPUs State
Time(s)
Domain-0 0 1890 1
r----- 0.0
mytestvm 149 60 1
-b---- 0.0
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel