2014-07-19 05:10:11

by Nicholas Krause

Subject: [PATCH] mips: Remove uneeded line in cmp_smp_finish

This patch removes a unneeded line from this file as stated by the
fix me in this file.

Signed-off-by: Nicholas Krause <[email protected]>
---
arch/mips/kernel/smp-cmp.c | 2 --
1 file changed, 2 deletions(-)

diff --git a/arch/mips/kernel/smp-cmp.c b/arch/mips/kernel/smp-cmp.c
index fc8a515..61bfa20 100644
--- a/arch/mips/kernel/smp-cmp.c
+++ b/arch/mips/kernel/smp-cmp.c
@@ -60,8 +60,6 @@ static void cmp_smp_finish(void)
{
pr_debug("SMPCMP: CPU%d: %s\n", smp_processor_id(), __func__);

- /* CDFIXME: remove this? */
- write_c0_compare(read_c0_count() + (8 * mips_hpt_frequency / HZ));

#ifdef CONFIG_MIPS_MT_FPAFF
/* If we have an FPU, enroll ourselves in the FPU-full mask */
--
1.9.1

2014-07-19 12:05:59

by Paul Bolle

Subject: Re: [PATCH] mips: Remove uneeded line in cmp_smp_finish

On Sat, 2014-07-19 at 01:10 -0400, Nicholas Krause wrote:
> This patch removes a unneeded line from this file as stated by the
> fix me in this file.
>
> Signed-off-by: Nicholas Krause <[email protected]>
> ---
> arch/mips/kernel/smp-cmp.c | 2 --
> 1 file changed, 2 deletions(-)
>
> diff --git a/arch/mips/kernel/smp-cmp.c b/arch/mips/kernel/smp-cmp.c
> index fc8a515..61bfa20 100644
> --- a/arch/mips/kernel/smp-cmp.c
> +++ b/arch/mips/kernel/smp-cmp.c
> @@ -60,8 +60,6 @@ static void cmp_smp_finish(void)
> {
> pr_debug("SMPCMP: CPU%d: %s\n", smp_processor_id(), __func__);
>
> - /* CDFIXME: remove this? */
> - write_c0_compare(read_c0_count() + (8 * mips_hpt_frequency / HZ));

That comment ends in a question mark. I wonder why...

> #ifdef CONFIG_MIPS_MT_FPAFF
> /* If we have an FPU, enroll ourselves in the FPU-full mask */


Paul Bolle

2014-07-19 21:33:18

by Nicholas Krause

Subject: Re: [PATCH] mips: Remove uneeded line in cmp_smp_finish

On Sat, Jul 19, 2014 at 8:05 AM, Paul Bolle <[email protected]> wrote:
> On Sat, 2014-07-19 at 01:10 -0400, Nicholas Krause wrote:
>> This patch removes a unneeded line from this file as stated by the
>> fix me in this file.
>>
>> Signed-off-by: Nicholas Krause <[email protected]>
>> ---
>> arch/mips/kernel/smp-cmp.c | 2 --
>> 1 file changed, 2 deletions(-)
>>
>> diff --git a/arch/mips/kernel/smp-cmp.c b/arch/mips/kernel/smp-cmp.c
>> index fc8a515..61bfa20 100644
>> --- a/arch/mips/kernel/smp-cmp.c
>> +++ b/arch/mips/kernel/smp-cmp.c
>> @@ -60,8 +60,6 @@ static void cmp_smp_finish(void)
>> {
>> pr_debug("SMPCMP: CPU%d: %s\n", smp_processor_id(), __func__);
>>
>> - /* CDFIXME: remove this? */
>> - write_c0_compare(read_c0_count() + (8 * mips_hpt_frequency / HZ));
>
> That comment ends in a question mark. I wonder why...
>
>> #ifdef CONFIG_MIPS_MT_FPAFF
>> /* If we have an FPU, enroll ourselves in the FPU-full mask */
>
>
> Paul Bolle
>
If we need it then can I remove the FIx me comment.
Cheers Nick

2014-07-21 08:19:39

by Paul Burton

Subject: Re: [PATCH] mips: Remove uneeded line in cmp_smp_finish

On Sat, Jul 19, 2014 at 05:33:16PM -0400, Nick Krause wrote:
> On Sat, Jul 19, 2014 at 8:05 AM, Paul Bolle <[email protected]> wrote:
> > On Sat, 2014-07-19 at 01:10 -0400, Nicholas Krause wrote:
> >> This patch removes a unneeded line from this file as stated by the
> >> fix me in this file.
> >>
> >> Signed-off-by: Nicholas Krause <[email protected]>
> >> ---
> >> arch/mips/kernel/smp-cmp.c | 2 --
> >> 1 file changed, 2 deletions(-)
> >>
> >> diff --git a/arch/mips/kernel/smp-cmp.c b/arch/mips/kernel/smp-cmp.c
> >> index fc8a515..61bfa20 100644
> >> --- a/arch/mips/kernel/smp-cmp.c
> >> +++ b/arch/mips/kernel/smp-cmp.c
> >> @@ -60,8 +60,6 @@ static void cmp_smp_finish(void)
> >> {
> >> pr_debug("SMPCMP: CPU%d: %s\n", smp_processor_id(), __func__);
> >>
> >> - /* CDFIXME: remove this? */
> >> - write_c0_compare(read_c0_count() + (8 * mips_hpt_frequency / HZ));
> >
> > That comment ends in a question mark. I wonder why...
> >
> >> #ifdef CONFIG_MIPS_MT_FPAFF
> >> /* If we have an FPU, enroll ourselves in the FPU-full mask */
> >
> >
> > Paul Bolle
> >
> If we need it then can I remove the FIx me comment.
> Cheers Nick

That depends: have you verified that we do need it?

I wouldn't feel comfortable with removing either line without someone
first explaining why it isn't necessary and testing the result on a
number of different systems, with varying combinations of csrc-r4k &
cevt-r4k.

I absolutely agree that removing unnecessary code or outdated comments
are both good things, but let's be sure they're unnecessary or outdated
first. Your patch does not make me confident that you've checked either
of those.

Thanks,
Paul

2014-07-21 17:34:57

by Nicholas Krause

Subject: Re: [PATCH] mips: Remove uneeded line in cmp_smp_finish

On Mon, Jul 21, 2014 at 4:19 AM, Paul Burton <[email protected]> wrote:
> On Sat, Jul 19, 2014 at 05:33:16PM -0400, Nick Krause wrote:
>> On Sat, Jul 19, 2014 at 8:05 AM, Paul Bolle <[email protected]> wrote:
>> > On Sat, 2014-07-19 at 01:10 -0400, Nicholas Krause wrote:
>> >> This patch removes a unneeded line from this file as stated by the
>> >> fix me in this file.
>> >>
>> >> Signed-off-by: Nicholas Krause <[email protected]>
>> >> ---
>> >> arch/mips/kernel/smp-cmp.c | 2 --
>> >> 1 file changed, 2 deletions(-)
>> >>
>> >> diff --git a/arch/mips/kernel/smp-cmp.c b/arch/mips/kernel/smp-cmp.c
>> >> index fc8a515..61bfa20 100644
>> >> --- a/arch/mips/kernel/smp-cmp.c
>> >> +++ b/arch/mips/kernel/smp-cmp.c
>> >> @@ -60,8 +60,6 @@ static void cmp_smp_finish(void)
>> >> {
>> >> pr_debug("SMPCMP: CPU%d: %s\n", smp_processor_id(), __func__);
>> >>
>> >> - /* CDFIXME: remove this? */
>> >> - write_c0_compare(read_c0_count() + (8 * mips_hpt_frequency / HZ));
>> >
>> > That comment ends in a question mark. I wonder why...
>> >
>> >> #ifdef CONFIG_MIPS_MT_FPAFF
>> >> /* If we have an FPU, enroll ourselves in the FPU-full mask */
>> >
>> >
>> > Paul Bolle
>> >
>> If we need it then can I remove the FIx me comment.
>> Cheers Nick
>
> That depends: have you verified that we do need it?
>
> I wouldn't feel comfortable with removing either line without someone
> first explaining why it isn't necessary and testing the result on a
> number of different systems, with varying combinations of csrc-r4k &
> cevt-r4k.
>
> I absolutely agree that removing unnecessary code or outdated comments
> are both good things, but let's be sure they're unnecessary or outdated
> first. Your patch does not make me confident that you've checked either
> of those.
>
> Thanks,
> Paul


I don't have this hardware , however if we can find people to test
this it would be great.
Cheers Nick