[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Xen-devel] [PATCH v2 07/12] xenctx: Add stack addr to call trace.

On 11/07/13 09:44, Ian Campbell wrote:
On Thu, 2013-11-07 at 09:34 -0500, Don Slutz wrote:
On 11/07/13 07:50, Ian Campbell wrote:

On Wed, 2013-11-06 at 15:08 -0500, Don Slutz wrote:
From: Don Slutz <Don@xxxxxxxxxxxxxxx>
Can you give an example of the output please.
Here it is (With all patches active):
           cs: 0010 @ 0000000000000000
Is the ugly wrapping real or email whitespace damage?
Real. The way it currently looks is clearer with a symbol. Doing both the address and symbol and other data is not simple. I was trying to keep things aligned which is very hard with variable length output. xen-hvmctx's output:

             cs 0x00000010 (0x0000000000000000 + 0xffffffff / 0x00a9b)

Using + to indicate a limit is not so clear either.

Things like idt are simpler.
     idt: ffffffff8042c000/fff idt_table

           ss: 0018 @ 0000000000000000
           ds: 0018 @ 0000000000000000
           es: 0018 @ 0000000000000000
           fs: 0000 @ 0000000000000000
           gs: 0000 @ ffffffff803ac000\0000000000000000 boot_cpu_pda\
Should this be split into "gs" and "shadow gs"?
What is "stack address" is it the base of the function's stack frame
perhaps? Or maybe the top? Or maybe the framepointer?
It is the address of the stack "word".
I'm afraid I am non the wiser.

Does just:

  ffffffff803ddf90: ffffffff80048d19 0000000000200800 .......... .....
  ffffffff803ddf90:  [<ffffffff80048d19>] cpu_idle+0x95

help?  The start of the line is the same.

Xen-devel mailing list



Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.