tech-kern archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

[PATCH] i386 copy.S routine optimization?



Hello,

I already talked about this with Radoslaw Kujawa on IRC, I understood that it is far from trivial to say if it is good to apply the following patch [0] or not due to x86 cache and pipeline subtleties.

I thought it was a good idea but I am indeed not an x86 expert at all :)

However, if anyone has an explanation (and the time to write it down) about why this might not be a good idea, I would be interested in reading it, just for my own personal curiosity.

If no one has the time, no problem :)

Best regards,

[0] -- https://github.com/fallen/NetBSD/commit/7080dc4a1f8861e5e3b8f4b0b05580698f1dc219

--
Yann Sionneau


Home | Main Index | Thread Index | Old Index