On Fri, Feb 24, 2017 at 06:11:30AM +0530, Nikunj Dadhania wrote: > On 24 February 2017 at 04:23, David Gibson <da...@gibson.dropbear.id.au> > wrote: > > On Fri, Feb 24, 2017 at 09:34:32AM +1100, Richard Henderson wrote: > >> On 02/23/2017 05:40 PM, Nikunj A Dadhania wrote: > >> > Richard Henderson <r...@twiddle.net> writes: > >> > > These functions are becoming quite large. Are they performance > >> > > critical enough > >> > > that they need to stay as inline code, or should they be moved to > >> > > helpers and > >> > > share code with cpu_read/write_xer? > >> > > >> > Just to boot to login prompt, these are the numbers for > >> > gen_read/write_xer: > >> > > >> > helper_myprint - rd_count 231103, wr_count 68897 > >> > > >> > And it keeps on incrementing, maybe scope of optimization here. > >> > >> That's not very large considering the total number of instructions executed > >> during a boot to prompt. > >> > >> Thoughts, David? > > > > Hm, I'm not clear if that's the number of executions, or the number of > > translations. > > That is number of executions.
Ok, I guess that's not that big, then. I guess moving them into
helpers would make sense.
Although I guess they'd shrink right down again if we put an
env->xer_mask in. Thoughts on that option Richard?
--
David Gibson | I'll have my music baroque, and my code
david AT gibson.dropbear.id.au | minimalist, thank you. NOT _the_ _other_
| _way_ _around_!
http://www.ozlabs.org/~dgibson
signature.asc
Description: PGP signature
