This is the mail archive of the libc-alpha@sourceware.org mailing list for the glibc project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: New optimized string routines for Intel and alignment of stack.


On 06/07/2016 01:51 PM, Adhemerval Zanella wrote:

Besides string routines, do we have any C code that relies on stack alignment
on 32-bit x86?

Perhaps not today, but unless we change the ABI to the original definition, GCC is free to exploit the stack alignment for optimization purposes in the future.

Also, is there any performance issue with current unaligned version or are you
just worry that we might remove them in the future due a better optimized version
that assumes aligned stack?

I think there is a performance penalty from not using vectorized copies for small structs. Even unaligned SSE loads/stores would be a win for the example I posted, I assume.

Thanks,
Florian


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]