Hi,
I've written a loop targeted at the C6600 using linear assembler. However I am puzzled by the outcome, although the optimizer states a schedule was found with 4 iterations in parallel, a non-piplined version is generated which is very slow of course.
In my previous attemps the optimizer stated the reason of not generating a pipelined loop, however in this case it remains silent.
Any ideas how to find out whats bothering the optimizer?
Thank you in advance, Clemens
[code];*----------------------------------------------------------------------------*
;* SOFTWARE PIPELINE INFORMATION
;*
;* Loop source line : 34
;* Loop closing brace source line : 131
;* Known Minimum Trip Count : 8
;* Known Max Trip Count Factor : 1
;* Loop Carried Dependency Bound(^) : 7
;* Unpartitioned Resource Bound : 11
;* Partitioned Resource Bound(*) : 11
;* Resource Partition:
;* A-side B-side
;* .L units 0 0
;* .S units 3 2
;* .D units 9 10
;* .M units 0 0
;* .X cross paths 2 2
;* .T address paths 9 10
;* Long read paths 0 0
;* Long write paths 0 0
;* Logical ops (.LS) 14 12 (.L or .S unit)
;* Addition ops (.LSD) 5 8 (.L or .S or .D unit)
;* Bound(.L .S .LS) 9 7
;* Bound(.L .S .D .LS .LSD) 11* 11*
;*
;* Searching for software pipeline schedule at ...
;* ii = 11 Schedule found with 4 iterations in parallel
;* Done
;*
;* Loop will be splooped
;* Collapsed epilog stages : 0
;* Collapsed prolog stages : 0
;* Minimum required memory pad : 0 bytes
;*
;* Minimum safe trip count : 1
;*----------------------------------------------------------------------------*[/code]