 |
Post AvvmW3MRCmva0IPHKC by [email protected] |
 |
More posts by [email protected] |
|
Post #AvtQffwaXMUuxvZxzs by [email protected] |
|
0 likes, 1 repeats |
|
|
|
fun kernel/compiler interaction that causes some Linux kernel code to have some… |
|
|
|
Post #AvtRCRyHVZENBc00kS by [email protected] |
|
0 likes, 1 repeats |
|
|
|
So in bar2, all of these instructions are unnecessary: pushq %rbp … |
|
|
|
Post #AvtRQ5sWRhIoijoME4 by [email protected] |
|
0 likes, 1 repeats |
|
|
|
(clang apparently does this better) |
|
|
|
Post #AvtSAacaTaaHOBWq9I by [email protected] |
|
0 likes, 0 repeats |
|
|
|
@jann huh, do you know why the kernel asks for only 8 byte alignment? |
|
|
|
Post #AvtSAaic7BP9gsLeXg by [email protected] |
|
0 likes, 1 repeats |
|
|
|
@osandov my understanding is that normal userspace code benefits from 16-byte s… |
|
|
|
Post #AvtdEpT8JYhk5lKYuu by [email protected] |
|
0 likes, 0 repeats |
|
|
|
@jann nasty. Apparently clang has an issue where the tail call optimization fai… |
|
|
|
Post #AvtdEpZruW5mQeTwPo by [email protected] |
|
0 likes, 1 repeats |
|
|
|
@irogers though if a stack allocation's address escapes the analysis, tail … |
|
|
|
Post #AvvhrLzWdKKM5z1k6i by [email protected] |
|
0 likes, 0 repeats |
|
|
|
@jann I'm pretty sure s1 is not a special case, but rather GCC upgrades ali… |
|
|
|
Post #AvvhrM6yBeHYT4Vgi8 by [email protected] |
|
0 likes, 0 repeats |
|
|
|
@jann want me to turn this into a GCC bugreport? |
|
|
|
Post #AvvhrMDLnvO0mrUmem by [email protected] |
|
0 likes, 1 repeats |
|
|
|
@amonakov oh, yes, thank you! |
|
|
|
Post #AvvmW3G3aVp7gVQBNY by [email protected] |
|
0 likes, 0 repeats |
|
|
|
@jann but it already works as desired if you pass -mgeneral-regs-only: https://… |
|
|
|
Post #AvvmW3MRCmva0IPHKC by [email protected] |
|
0 likes, 1 repeats |
|
|
|
@amonakov hmm... x86 Linux doesn't use -fgeneral-regs-only, but it does KBU… |
