What to do with gcc 3.4 regression on Itanium 1?

[Vladimir N. Makarov]

H. J. Lu wrote:

>On Tue, Jun 01, 2004 at 12:27:09AM -0700, Jim Wilson wrote:
>  
>
>>H. J. Lu wrote:
>>    
>>
>>>It looks like a typo to me.
>>>      
>>>
>>Which part exactly is the typo?  These ambiguous bug reports are not 
>>very useful.
>>
>>It looks OK to me.  1_ialu is used when both bundling_p and 
>>ia64_produce_address_p are false.  1_ialu_addr is used when bundling is 
>>false and ia64_produce_address_p is true.  1b_ialu is used when bundling 
>>is true and ia64_produce_address_p is false.  1b_ialu_addr is used when 
>>both bundling are ia64_produce_address_p are true.
>>    
>>
>
>Doesn't "1_M"/"1b_M" mean that "1_ialu_addr"/"1b_ialu_addr" will be
>used for M unit only? That contradicts the comment
>
>;; ??? howto describe ialu for I slot only.  We use ialu_addr for that
>
>which implies "ialu_addr" is for I unit.
>
>  
>
I've made two patches on Monday to solve the problems but because of the 
gcc sunnit I had no opportunity to send it. I'll do it tomorrow. One 
problem is fxied in bundling (and could be considered as typo). Another 
one is more serious and will be fixed in insn-scheduler (the reason of 
the problem is discrepancy of work of ia64 bundler and insn scheduler. 
Insn scheduler never updates DFA state for asm-insn but the bundler does).

Vlad

>  
>