Hi,
I just got CCS 5.2 work on my PC. I run the optimization tutorial (tutor_d.c, lesson_c.c, lesson1_c.c, lesson2_c.c and lesson3_c.c from CCS 3.1) on CCS 5.2. I find that the v7.3.1 compiling results are different from v.5.3.1 of CCS 3.1.
v7.3.1:
;*********** ;* TMS320C6x C/C++ Codegen PC v7.3.1 * ;* Date/Time created: Thu Jun 14 11:01:16 2012 *
;*********** .compiler_opts --abi=coffabi --c64p_l1d_workaround=default --endian=little --hll_source=on --long_precision_bits=40 --mem_model:code=near --mem_model:const=data --mem_model:data=far_aggregates --object_format=coff --silicon_version=6200 --symdebug:dwarf
;******* ;* GLOBAL FILE PARAMETERS *
;* Architecture : TMS320C620x * ;* Optimization : Enabled at level 3 * ;* Optimizing for : Speed * ;* Based on options: -o3, no -ms * ;* Endian : Little * ;* Interrupt Thrshld : Disabled * ;* Data Access Model : Far Aggregate Data * ;* Pipelining : Enabled * ;* Speculate Loads : Enabled with threshold = 0 * ;* Memory Aliases : Presume are aliases (pessimistic) * ;* Debug Info : DWARF Debug * ;* * ;*******
v5.3.1:
;* TMS320C6x C/C++ Codegen PC v5.1.0 * ;* Date/Time created: Wed Jun 13 03:25:28 2012 *
;******** .compiler_opts --endian=little --mem_model:code=near --mem_model:data=far_aggregates --quiet --silicon_version=6210
;********** ;* GLOBAL FILE PARAMETERS * ;* *
;* Architecture : TMS320C621x * ;* Optimization : Enabled at level 3 * ;* Optimizing for : Speed * ;* Based on options: -o3, no -ms * ;* Endian : Little * ;* Interrupt Thrshld : Disabled * ;* Data Access Model : Far Aggregate Data * ;* Pipelining : Enabled * ;* Speculate Loads : Disabled * ;* Memory Aliases : Presume are aliases (pessimistic) *
;* Debug Info : DWARF Debug *
BTW, I cannot change the architecture in CCS 5.2 from TMS320C620x to TMS320C621x as that of CCS3.1. I do not find any option can change that.
The more important thing is the software pipeline result differences:
v7.3.1:
;* SOFTWARE PIPELINE INFORMATION ;* ;* Loop source line : 35 ;* Loop opening brace source line : 36 ;* Loop closing brace source line : 40 ;* Loop Unroll Multiple : 2x ;* Known Minimum Trip Count : 5 ;* Known Max Trip Count Factor : 1 ;* Loop Carried Dependency Bound(^) : 0 ;* Unpartitioned Resource Bound : 3 ;* Partitioned Resource Bound(*) : 3 ;* Resource Partition: ;* A-side B-side ;* .L units 0 0 ;* .S units 2 1 ;* .D units 3* 3* ;* .M units 2 2 ;* .X cross paths 1 1 ;* .T address paths 3* 3* ;* Long read paths 1 1 ;* Long write paths 0 0 ;* Logical ops (.LS) 1 1 (.L or .S unit) ;* Addition ops (.LSD) 0 1 (.L or .S or .D unit) ;* Bound(.L .S .LS) 2 1 ;* Bound(.L .S .D .LS .LSD) 2 2 ;* ;* Searching for software pipeline schedule at ... ;* ii = 3 Schedule found with 5 iterations in parallel ;* Done ;* ;* Epilog not removed ;* Collapsed epilog stages : 0 ;* ;* Prolog not entirely removed ;* Collapsed prolog stages : 3 ;* ;* Minimum required memory pad : 0 bytes ;* ;* For further improvement on this loop, try option -mh16 ;* ;* Minimum safe trip count : 4 (after unrolling)
v5.3.1:
;* SOFTWARE PIPELINE INFORMATION ;* ;* Loop source line : 35 ;* Loop opening brace source line : 36 ;* Loop closing brace source line : 40 ;* Loop Unroll Multiple : 2x ;* Known Minimum Trip Count : 5 ;* Known Max Trip Count Factor : 1 ;* Loop Carried Dependency Bound(^) : 1 ;* Unpartitioned Resource Bound : 2 ;* Partitioned Resource Bound(*) : 2 ;* Resource Partition: ;* A-side B-side ;* .L units 0 0 ;* .S units 2* 1 ;* .D units 2* 2* ;* .M units 2* 2* ;* .X cross paths 1 1 ;* .T address paths 2* 2* ;* Long read paths 1 1 ;* Long write paths 0 0 ;* Logical ops (.LS) 1 1 (.L or .S unit) ;* Addition ops (.LSD) 0 1 (.L or .S or .D unit) ;* Bound(.L .S .LS) 2* 1 ;* Bound(.L .S .D .LS .LSD) 2* 2* ;* ;* Searching for software pipeline schedule at ... ;* ii = 2 Schedule found with 6 iterations in parallel ;* ;* Register Usage Table: ;* +---------------------------------+ ;* |AAAAAAAAAAAAAAAA|BBBBBBBBBBBBBBBB| ;* |0000000000111111|0000000000111111| ;* |0123456789012345|0123456789012345| ;* |----------------+----------------| ;* 0: | * ******* |*** ****** | ;* 1: | * ******* |*** ****** | ;* +---------------------------------+ ;* ;* Done ;* ;* Epilog not entirely removed ;* Collapsed epilog stages : 2 ;* ;* Prolog not entirely removed ;* Collapsed prolog stages : 3 ;* ;* Minimum required memory pad : 0 bytes ;* ;* For further improvement on this loop, try option -mh20 ;* ;* Minimum safe trip count : 3 (after unrolling)
........................
It seems v7.3.1 result is not as good as that of v5.3.1. On CCS5.2, I do not find where to change to -mh16 as CCS5.2 recommands. Is my result true? Or, I miss something? Thanks in advance.
