Re: timecounter changes break netbsd-9/i386 for NET4501

[SAITOH Masanobu <msaitoh%execsw.org@localhost>]

Hi, John.

On 2020/07/27 4:33, John D. Baker wrote:
> The following patch fixes the problem:
>
> +Index: sys/arch/x86/x86/cpu.c
> +===================================================================
> +RCS file: /cvsroot/src/sys/arch/x86/x86/cpu.c,v
> +retrieving revision 1.171.2.2
> +diff -u -p -r1.171.2.2 cpu.c
> +--- sys/arch/x86/x86/cpu.c	15 Jul 2020 17:25:08 -0000	1.171.2.2
> ++++ sys/arch/x86/x86/cpu.c	26 Jul 2020 17:30:27 -0000
> +@@ -1267,7 +1267,7 @@ cpu_get_tsc_freq(struct cpu_info *ci)
> + {
> + 	uint64_t freq = 0, last_tsc;
> +
> +-	if (cpu_hascounter())
> ++	if (cpu_hascounter()) {
> + 		freq = cpu_tsc_freq_cpuid(ci);
> +
> + 	if (freq != 0) {
> +@@ -1280,6 +1280,7 @@ cpu_get_tsc_freq(struct cpu_info *ci)
> + 		ci->ci_data.cpu_cc_freq =
> + 		    (cpu_counter_serializing() - last_tsc) * 10;
> + 	}
> ++	}
> + }
> +
> + void

Yes, your patch is correct.

I'll send new pullup request to fix this problem soon.

I'm sorry for the regression and thank you very much.

--
-----------------------------------------------
                SAITOH Masanobu (msaitoh%execsw.org@localhost
                                 msaitoh%netbsd.org@localhost)