[prev in list] [next in list] [prev in thread] [next in thread] 

List:       xen-ia64-devel
Subject:    RE: [Xen-ia64-devel] Community effort needed
From:       "Tian, Kevin" <kevin.tian () intel ! com>
Date:       2005-08-31 14:29:13
Message-ID: 571ACEFD467F7749BC50E0A98C17CDD807C324D3 () pdsmsx403
[Download RAW message or body]

Hi, John and Dan,
	Attached are the latest patches to promise domU running to
following point and however MCA then. Seems the possible cause is the
update of grant table code. You can continue debug based upon them.

	Currently I can observe many successful transactions between
blkback and blkfront, and also xenstore data already shows some
necessary relationship between them. To allocate an extra xenstore mfn
in control panel, I have to hack temporarily for get_domaininfo to
return max_pages as tot_pages. The reason is that currently ia64 control
panel uses an allocate-on-demand style to allocate domU memory, which
differ with x86 which allocate all necessary machine pages for domU at
creation. So tot_pages on ia64 can't satisfy the assumption in other
common code, which makes us really difficult to adopt new features
easily like this one. After making current repository working, I'll push
out a patch to cleanup those variances.

	Due to some conceptual change in grant table code, (like
dev_bus_addr becomes a real va, instead of frame number), so I pulled
common/grant_table.c into arch/ia64 with all previous ia64 specific
changes keeped. Plus some fix in xenolinux code, I observed some mapping
requests returned successfully. But no time to validate the content yet,
and this change may bring some implicit errors...

	Normally when communication between blkback and blkfront is
broken, there's warning about "timeout to connect..." just after "After
init_blk_xenbus". But now it's missing which just mean something good.
Anyway, now it should be very close to the end, and hope to see good
news from you guys stomorrow. ;-)

Thanks,
Kevin

(XEN) After init_blk_xenbus
(XEN) NET: Registered protocol family 2
(XEN) rs/usb/input/hid-core.c: v2.01:USB HID core driver
mice: PS/2 mouse device common for all mice
EFI Variables Facility v0.08 20
IP: routing cache hash table of 4096 buckets, 64Kbytes
(XEN) TCP established hash table entries: 65536 (order: 6, 1048576
bytes)
(XEN) TCP bind hash table entries: 65536 (order: 6, 1048576 bytes)
(XEN) rs: get_next_variable: status=80
(XEN) TCP: Hash tables configured (established 65536 bind 65536)
(XEN) NET: Registered protocol family 1
(XEN) NET: Registered protocol family 17

translate_domain_pte: bad mpa=000001000000000c (>
0000000010004000),vadr=e00001000000000c,pteval=0010010000000661,itir=000
0000000000760
grant table at e000010000000000
after gnttab_resume 
(XEN) translate_domain_pte: bad mpa=000001000000000c (>
0000000010004000),vadr=e00001000000000c,pteval=0010010000000661,itir=000
0000000000760
Grant tabl
(MCA reset happens then!!!)

["hg_cleanup_0831_xeno" (application/octet-stream)]
["hg_cleanup_0831_libxc" (application/octet-stream)]
["hg_cleanup_0831_xen" (application/octet-stream)]

_______________________________________________
Xen-ia64-devel mailing list
Xen-ia64-devel@lists.xensource.com
http://lists.xensource.com/xen-ia64-devel

[prev in list] [next in list] [prev in thread] [next in thread]