2010/5/28 David Daney <david.s.daney@xxxxxxxxx>: > This depends on not consistent with the #if conditions in [01/12] for pmu.h. > They should be I think. [DC]: It's a subset. If the code passes the test on other CPUs (such as CPU_SB1), we can add them here. > Probably removing the tests from pmu.h and encoding them here is better. [DC]: According to my comments for [1/12], how about keeping them untouched for now? > IANAL, but who holds the copyright? You or MTI ? [DC]: OK. Will change like this: Copyright (C) 2010 MIPS Technologies, Inc. Author: Deng-Cheng Zhu > This shadows the declaration in asm/time.h. Declare it in exactly one place > please. [DC]: OK. > Same thing about the copyright. [DC]: OK. > Not quite true. They use the high bit, that can be either 31 or 63 > depending on the width of the counters. [DC]: OK. Will change to consider 64-bit counters. > Counters can be 64-bits wide, unsigned int is only 32-bits wide. [DC]: OK. Deng-Cheng
