mirror of
https://github.com/c64scene-ar/llvm-6502.git
synced 2024-12-24 06:30:19 +00:00
add a note
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@122559 91177308-0d34-0410-b5e6-96231b3b80d8
This commit is contained in:
parent
7ab3cc32d6
commit
8a4893e013
@ -96,6 +96,97 @@ It appears icc use push for parameter passing. Need to investigate.
|
||||
|
||||
//===---------------------------------------------------------------------===//
|
||||
|
||||
This:
|
||||
|
||||
void foo(void);
|
||||
void bar(int x, int *P) {
|
||||
x >>= 2;
|
||||
if (x)
|
||||
foo();
|
||||
*P = x;
|
||||
}
|
||||
|
||||
compiles into:
|
||||
|
||||
movq %rsi, %rbx
|
||||
movl %edi, %r14d
|
||||
sarl $2, %r14d
|
||||
testl %r14d, %r14d
|
||||
je LBB0_2
|
||||
|
||||
Instead of doing an explicit test, we can use the flags off the sar. This
|
||||
occurs in a bigger testcase like this, which is pretty common:
|
||||
|
||||
#include <vector>
|
||||
|
||||
|
||||
int test1(std::vector<int> &X) {
|
||||
int Sum = 0;
|
||||
for (long i = 0, e = X.size(); i != e; ++i)
|
||||
X[i] = 0;
|
||||
return Sum;
|
||||
}
|
||||
compiles into:
|
||||
|
||||
movq %rsi, %rbx
|
||||
movl %edi, %r14d
|
||||
sarl $2, %r14d
|
||||
testl %r14d, %r14d
|
||||
je LBB0_2
|
||||
|
||||
Instead of doing an explicit test, we can use the flags off the sar. This
|
||||
occurs in a bigger testcase like this, which is pretty common:
|
||||
|
||||
#include <vector>
|
||||
|
||||
|
||||
int test1(std::vector<int> &X) {
|
||||
int Sum = 0;
|
||||
for (long i = 0, e = X.size(); i != e; ++i)
|
||||
X[i] = 0;
|
||||
return Sum;
|
||||
}
|
||||
compiles into:
|
||||
|
||||
movq %rsi, %rbx
|
||||
movl %edi, %r14d
|
||||
sarl $2, %r14d
|
||||
testl %r14d, %r14d
|
||||
je LBB0_2
|
||||
|
||||
Instead of doing an explicit test, we can use the flags off the sar. This
|
||||
occurs in a bigger testcase like this, which is pretty common:
|
||||
|
||||
#include <vector>
|
||||
|
||||
|
||||
int test1(std::vector<int> &X) {
|
||||
int Sum = 0;
|
||||
for (long i = 0, e = X.size(); i != e; ++i)
|
||||
X[i] = 0;
|
||||
return Sum;
|
||||
}
|
||||
compiles into:
|
||||
|
||||
movq %rsi, %rbx
|
||||
movl %edi, %r14d
|
||||
sarl $2, %r14d
|
||||
testl %r14d, %r14d
|
||||
je LBB0_2
|
||||
|
||||
Instead of doing an explicit test, we can use the flags off the sar. This
|
||||
occurs in a bigger testcase like this, which is pretty common in bootstrap:
|
||||
|
||||
#include <vector>
|
||||
int test1(std::vector<int> &X) {
|
||||
int Sum = 0;
|
||||
for (long i = 0, e = X.size(); i != e; ++i)
|
||||
X[i] = 0;
|
||||
return Sum;
|
||||
}
|
||||
|
||||
//===---------------------------------------------------------------------===//
|
||||
|
||||
Only use inc/neg/not instructions on processors where they are faster than
|
||||
add/sub/xor. They are slower on the P4 due to only updating some processor
|
||||
flags.
|
||||
|
Loading…
Reference in New Issue
Block a user