svn commit: r224187 - in head: sys/amd64/amd64 sys/arm/arm sys/arm/sa11x0 sys/i386/i386 sys/ia64/ia64 sys/kern sys/mips/mips sys/powerpc/aim sys/powerpc/booke sys/sparc64/sparc64 sys/sys usr.bin/vm...

[Marcel Moolenaar]

On Jul 18, 2011, at 5:59 PM, Attilio Rao wrote:

> 2011/7/19 Marcel Moolenaar <marcel at xcllnt.net>:
>> 
>> On Jul 18, 2011, at 8:19 AM, Attilio Rao wrote:
>> 
>>> Author: attilio
>>> Date: Mon Jul 18 15:19:40 2011
>>> New Revision: 224187
>>> URL: http://svn.freebsd.org/changeset/base/224187
>>> 
>>> Log:
>>>  - Remove the eintrcnt/eintrnames usage and introduce the concept of
>>>    sintrcnt/sintrnames which are symbols containing the size of the 2
>>>    tables.
>>>  - For amd64/i386 remove the storage of intr* stuff from assembly files.
>>>    This area can be widely improved by applying the same to other
>>>    architectures and likely finding an unified approach among them and
>>>    move the whole code to be MI. More work in this area is expected to
>>>    happen fairly soon.
>>> 
>>>  No MFC is previewed for this patch.
>> 
>> You just broke ia64 and possibly other 64-bit architectures:
>> 
>> ".word" declares a 16-bit integral on ia64 and the size symbols
>> are of type size_t (=64 bit). We'll be having misaligned loads
>> (= kernel panics) and/or reading garbage...
> 
> I'm a bit surprised of this though.
> .hword was supposed to be the 16-bit integral, while .word was
> supposed to be the 32-bits one, if I read my "info as" on amd64.

Well... all I can say is that assembly is the least transposable
language, besides of course machine code itself :-)

> Anyway, what do you think about this patch? (I still need to test it):
> http://www.freebsd.org/~attilio/64bits-fixup.diff

Looks good to me, though I don't know enough about mips to comment
on that. I'm not going to be anal about the use of ".quad" instead
of "data8" for ia64 -- let's get it fixed first (I think we have
".byte" in locore.S anyway :-)

-- 
Marcel Moolenaar
marcel at xcllnt.net