Abhinash Jain
2013-Sep-22 04:17 UTC
[LLVMdev] Why total number of store instructions increased in case of gvn with reg2mem?
Below are the 4 cases out of which for the first 3 cases am able to understand the variation on total number of store inst. But when -gvn parameters is passed (in case 4) , why does the total number of strore Instruction present on the attached file increases (instead of decrease). Please explain considering that use of -reg2mem pass is mandatory. eg. clang -emit-llvm bzip.c -c -o bzip1.bc 1. opt -load /home/honey/llvm/build/Debug+Asserts/lib/loadInstruction.so -storeInstruction < bzip1.bc output :- Total Store Instruction 3115 2. opt -reg2mem -load /home/honey/llvm/build/Debug+Asserts/lib/loadInstruction.so -storeInstruction < bzip1.bc output :- Total Store Instruction 3174 3. opt -dse -dce -reg2mem -load /home/xxx/llvm/build/Debug+Asserts/lib/loadInstruction.so -storeInstruction < bzip1.bc output :- Total Store Instruction 3171 4. opt -gvn -dse -dce -reg2mem -load /home/xxx/llvm/build/Debug+Asserts/lib/loadInstruction.so -storeInstruction < bzip1.bc output :- Total Store Instruction 4489 bzip.c <http://llvm.1065342.n5.nabble.com/file/n61412/bzip.c> -- View this message in context: http://llvm.1065342.n5.nabble.com/Why-total-number-of-store-instructions-increased-in-case-of-gvn-with-reg2mem-tp61412.html Sent from the LLVM - Dev mailing list archive at Nabble.com.
Seemingly Similar Threads
- [LLVMdev] Why total number of store instructions increased in case of gvn with reg2mem?
- [LLVMdev] Inneffiecient code produced by reg2mem?
- [LLVMdev] eliminate phi nodes, reduce unnecessary loads / stores , reg2mem, mem2reg
- [LLVMdev] run -mem2reg and -reg2mem programmably from within a Pass
- [LLVMdev] Reg2mem: Identifying introduced memory locations (also, what happens to the phi nodes)