cvs commit: src/lib/libc/sparc64 Symbol.map src/sys/sparc64/include profile.h src/sys/sun4v/include profile.h

[Robert Watson]

On Sat, 12 May 2007, Kris Kennaway wrote:

> On Sat, May 12, 2007 at 12:16:51PM +0200, Marius Strobl wrote:
>> On Fri, May 11, 2007 at 10:22:21PM -0400, Kris Kennaway wrote:
>>> On Fri, May 11, 2007 at 11:43:55PM +0000, Marius Strobl wrote:
>>>> marius      2007-05-11 23:43:55 UTC
>>>>
>>>>   FreeBSD src repository
>>>>
>>>>   Modified files:
>>>>     lib/libc/sparc64     Symbol.map
>>>>     sys/sparc64/include  profile.h
>>>>     sys/sun4v/include    profile.h
>>>>   Log:
>>>>   - Add bits for userland profiling. For sun4u this is compile-tested 
>>>> only.
>>>
>>> Does this mean you're able to run sun4v?  When I checked the other week it 
>>> was broken by the interrupt filter changes back in February.
>>
>> Ah, no, that means I made a typo; for sun4v it's compile-tested only :)
>
> OK, probably a followup commit would be good.  Inaccurate comments in commit 
> logs have a way of confusing people when they look at them years later.

Speaking of sun4v -- it sounds like sun4v support in 7.0 is at risk.  I think 
I understand from a comment I saw from Kip that there are (at least) two 
issues:

(1) An issue in VM handling that will be resolved by alc's replacement
     contigmalloc.  I don't know what the details are here.

(2) An issue involving sun4v breakage post-ifilters.  No details here either.

If we're going to ship a 7.0 release for sun4v, we need to get these things 
resolved.  Do we have a sense of what work is required here?

Robert N M Watson
Computer Laboratory
University of Cambridge