[PATCH] floppy: revert floppy disable halt warning

[PATCH] floppy: revert floppy disable halt warning

[Stephen Hemminger]

Remove the stupid floppy disable halt warning. It is meaningless since
the API is local to floppy driver.  Our boot scripts check for the
floppy drive, and this causes the warning to trigger every time. This
confuses support, breaks automated tests that look for backtraces on
boot,  and adds no value.  

The message is being displayed to the wrong audience, it looks like it
was intended as a compromise of a long discussion on LKML about
supporting older x86 hardware; but users don't know or understand what
it is saying.

If you want to change halt handling then just fix
the floppy driver, don't whine about it.

Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>

--- a/drivers/block/floppy.c	2012-03-11 09:27:52.866459327 -0700
+++ b/drivers/block/floppy.c	2012-03-11 09:28:29.286579201 -0700
@@ -1037,7 +1037,6 @@ static void floppy_disable_hlt(void)
 {
 	unsigned long flags;
 
-	WARN_ONCE(1, "floppy_disable_hlt() scheduled for removal in 2012");
 	spin_lock_irqsave(&floppy_hlt_lock, flags);
 	if (!hlt_disabled) {
 		hlt_disabled = 1;

Re: [PATCH] floppy: revert floppy disable halt warning

[Andrew Morton]

(cc some x86 greybeards)

On Sun, 11 Mar 2012 09:38:53 -0700
Stephen Hemminger <shemminger@vyatta.com> wrote:

> Remove the stupid floppy disable halt warning. It is meaningless since
> the API is local to floppy driver.  Our boot scripts check for the
> floppy drive, and this causes the warning to trigger every time. This
> confuses support, breaks automated tests that look for backtraces on
> boot,  and adds no value.  
> 
> The message is being displayed to the wrong audience, it looks like it
> was intended as a compromise of a long discussion on LKML about
> supporting older x86 hardware; but users don't know or understand what
> it is saying.
> 
> If you want to change halt handling then just fix
> the floppy driver, don't whine about it.
> 
> ...
> 
> --- a/drivers/block/floppy.c	2012-03-11 09:27:52.866459327 -0700
> +++ b/drivers/block/floppy.c	2012-03-11 09:28:29.286579201 -0700
> @@ -1037,7 +1037,6 @@ static void floppy_disable_hlt(void)
>  {
>  	unsigned long flags;
>  
> -	WARN_ONCE(1, "floppy_disable_hlt() scheduled for removal in 2012");
>  	spin_lock_irqsave(&floppy_hlt_lock, flags);
>  	if (!hlt_disabled) {
>  		hlt_disabled = 1;

It would have been nice to have provided a pointer to this "long
discussion on LKML".  I tried for a while, then gave up.

In my search I came across
http://lkml.indiana.edu/hypermail/linux/kernel/1104.0/00461.html which
does indeed remove all the disable_hlt() code as well as the warning
and the feature-removal-schedule.txt record.  Why didn't we merge that?
Why shouldn't we merge that now?

Re: [PATCH] floppy: revert floppy disable halt warning

[Josh Boyer]

On Tue, Mar 27, 2012 at 7:26 PM, Andrew Morton
<akpm@linux-foundation.org> wrote:
> (cc some x86 greybeards)
>
> On Sun, 11 Mar 2012 09:38:53 -0700
> Stephen Hemminger <shemminger@vyatta.com> wrote:
>
>> Remove the stupid floppy disable halt warning. It is meaningless since
>> the API is local to floppy driver.  Our boot scripts check for the
>> floppy drive, and this causes the warning to trigger every time. This
>> confuses support, breaks automated tests that look for backtraces on
>> boot,  and adds no value.
>>
>> The message is being displayed to the wrong audience, it looks like it
>> was intended as a compromise of a long discussion on LKML about
>> supporting older x86 hardware; but users don't know or understand what
>> it is saying.
>>
>> If you want to change halt handling then just fix
>> the floppy driver, don't whine about it.
>>
>> ...
>>
>> --- a/drivers/block/floppy.c  2012-03-11 09:27:52.866459327 -0700
>> +++ b/drivers/block/floppy.c  2012-03-11 09:28:29.286579201 -0700
>> @@ -1037,7 +1037,6 @@ static void floppy_disable_hlt(void)
>>  {
>>       unsigned long flags;
>>
>> -     WARN_ONCE(1, "floppy_disable_hlt() scheduled for removal in 2012");
>>       spin_lock_irqsave(&floppy_hlt_lock, flags);
>>       if (!hlt_disabled) {
>>               hlt_disabled = 1;
>
> It would have been nice to have provided a pointer to this "long
> discussion on LKML".  I tried for a while, then gave up.
>
> In my search I came across
> http://lkml.indiana.edu/hypermail/linux/kernel/1104.0/00461.html which
> does indeed remove all the disable_hlt() code as well as the warning
> and the feature-removal-schedule.txt record.  Why didn't we merge that?
> Why shouldn't we merge that now?

And again 5 months later:

https://lkml.org/lkml/2011/10/28/313

Third time's a charm?  I'd buy someone a beverage if we actually went
through our planned feature removal here...

josh

Re: [PATCH] floppy: revert floppy disable halt warning

[Ingo Molnar]

* Josh Boyer <jwboyer@gmail.com> wrote:

> On Tue, Mar 27, 2012 at 7:26 PM, Andrew Morton
> <akpm@linux-foundation.org> wrote:
> > (cc some x86 greybeards)
> >
> > On Sun, 11 Mar 2012 09:38:53 -0700
> > Stephen Hemminger <shemminger@vyatta.com> wrote:
> >
> >> Remove the stupid floppy disable halt warning. It is meaningless since
> >> the API is local to floppy driver.  Our boot scripts check for the
> >> floppy drive, and this causes the warning to trigger every time. This
> >> confuses support, breaks automated tests that look for backtraces on
> >> boot,  and adds no value.
> >>
> >> The message is being displayed to the wrong audience, it looks like it
> >> was intended as a compromise of a long discussion on LKML about
> >> supporting older x86 hardware; but users don't know or understand what
> >> it is saying.
> >>
> >> If you want to change halt handling then just fix
> >> the floppy driver, don't whine about it.
> >>
> >> ...
> >>
> >> --- a/drivers/block/floppy.c  2012-03-11 09:27:52.866459327 -0700
> >> +++ b/drivers/block/floppy.c  2012-03-11 09:28:29.286579201 -0700
> >> @@ -1037,7 +1037,6 @@ static void floppy_disable_hlt(void)
> >>  {
> >>       unsigned long flags;
> >>
> >> -     WARN_ONCE(1, "floppy_disable_hlt() scheduled for removal in 2012");
> >>       spin_lock_irqsave(&floppy_hlt_lock, flags);
> >>       if (!hlt_disabled) {
> >>               hlt_disabled = 1;
> >
> > It would have been nice to have provided a pointer to this "long
> > discussion on LKML".  I tried for a while, then gave up.
> >
> > In my search I came across
> > http://lkml.indiana.edu/hypermail/linux/kernel/1104.0/00461.html which
> > does indeed remove all the disable_hlt() code as well as the warning
> > and the feature-removal-schedule.txt record.  Why didn't we merge that?
> > Why shouldn't we merge that now?
> 
> And again 5 months later:
> 
> https://lkml.org/lkml/2011/10/28/313
> 
> Third time's a charm?  I'd buy someone a beverage if we actually went
> through our planned feature removal here...

No objections from me.

Please resubmit against latest -tip, we unified the idle 
routines which likely interacts with that patch.

Thanks,

	Ingo

Re: [PATCH] floppy: revert floppy disable halt warning

[H. Peter Anvin]

On 03/27/2012 11:40 PM, Ingo Molnar wrote:
> 
> No objections from me.
> 
> Please resubmit against latest -tip, we unified the idle 
> routines which likely interacts with that patch.
> 

To the best of my knowledge, no-hlt only existed because of flaky power
distributions on 386/486 systems which were sold to run DOS.  Since DOS
did no power management of any kind, including HLT, the power draw was
fairly uniform; when exposed to the much higher noise levels you got
when Linux used HLT caused some of these systems to fail.

They were by far in the minority even back then.

	-hpa

-- 
H. Peter Anvin, Intel Open Source Technology Center
I work for Intel.  I don't speak on their behalf.

Re: [PATCH] floppy: revert floppy disable halt warning

[Ingo Molnar]

* H. Peter Anvin <hpa@zytor.com> wrote:

> On 03/27/2012 11:40 PM, Ingo Molnar wrote:
> > 
> > No objections from me.
> > 
> > Please resubmit against latest -tip, we unified the idle 
> > routines which likely interacts with that patch.
> 
> To the best of my knowledge, no-hlt only existed because of 
> flaky power distributions on 386/486 systems which were sold 
> to run DOS.  Since DOS did no power management of any kind, 
> including HLT, the power draw was fairly uniform; when exposed 
> to the much higher noise levels you got when Linux used HLT 
> caused some of these systems to fail.
> 
> They were by far in the minority even back then.

This information should be added to the changelog.

Thanks,

	Ingo

Re: [PATCH] floppy: revert floppy disable halt warning

[Alan Cox]

On Wed, 28 Mar 2012 00:20:46 -0700
"H. Peter Anvin" <hpa@zytor.com> wrote:

> On 03/27/2012 11:40 PM, Ingo Molnar wrote:
> > 
> > No objections from me.
> > 
> > Please resubmit against latest -tip, we unified the idle 
> > routines which likely interacts with that patch.
> > 
> 
> To the best of my knowledge, no-hlt only existed because of flaky power
> distributions on 386/486 systems which were sold to run DOS.  Since DOS
> did no power management of any kind, including HLT, the power draw was
> fairly uniform; when exposed to the much higher noise levels you got
> when Linux used HLT caused some of these systems to fail.

Also for the Cyrix 5510 which tended to go castors up if a HLT occurred
during a DMA cycle and on a few other boxes HLT during DMA tended to go
astray.

Do we care ? I doubt it. The 5510 was pretty obscure, the 5520 fixed it,
the 5530 is probably the oldest still in any kind of use.

Alan