Re: CVS commit: src/sys/arch

[Maxime Villard]

Le 04/12/2017 à 20:17, Christos Zoulas a écrit :

In article ,
Maxime Villard   wrote:

Le 03/12/2017 à 16:55, David Holland a écrit :

On Sun, Dec 03, 2017 at 10:05:08AM +0100, Maxime Villard wrote:
   > Good, and apparently no one has noticed that some options don't compile.

Are you saying Christos is lying about it working after reverting your
commits?


Why don't you read the previous emails from this thread, instead of coming
up with random, inaccurate statements like this one? I said very clearly that
COMPAT_20 for example does not compile. Same for COMPAT_16 and COMPAT_10,
actually.


By "does not compile" you probably mean "does not compile for me". Can you
please post your config file? Perhaps it is easy to fix.


Give me a few days and I'll handle that (along with reverting the things).

Re: CVS commit: src/sys/arch

[Christos Zoulas]

In article ,
Maxime Villard   wrote:
>Le 03/12/2017 à 16:55, David Holland a écrit :
>> On Sun, Dec 03, 2017 at 10:05:08AM +0100, Maxime Villard wrote:
>>   > Good, and apparently no one has noticed that some options don't compile.
>> 
>> Are you saying Christos is lying about it working after reverting your
>> commits?
>
>Why don't you read the previous emails from this thread, instead of coming
>up with random, inaccurate statements like this one? I said very clearly that
>COMPAT_20 for example does not compile. Same for COMPAT_16 and COMPAT_10,
>actually.

By "does not compile" you probably mean "does not compile for me". Can you
please post your config file? Perhaps it is easy to fix.

christos

Re: CVS commit: src/sys/arch

[Maxime Villard]

Le 03/12/2017 à 16:55, David Holland a écrit :

On Sun, Dec 03, 2017 at 10:05:08AM +0100, Maxime Villard wrote:
  > Good, and apparently no one has noticed that some options don't compile.

Are you saying Christos is lying about it working after reverting your
commits?


Why don't you read the previous emails from this thread, instead of coming
up with random, inaccurate statements like this one? I said very clearly that
COMPAT_20 for example does not compile. Same for COMPAT_16 and COMPAT_10,
actually.


  > As I said earlier, this "please ask core for approval" argument
  > does not work on me anymore, sorry about that. So I'm going to
  > commit right away now, on whatever directly has to do with amd64.

"sorry about that" is not the right answer. Like it or not, [...] drama etc


See the internal thread.

Maxime

Re: CVS commit: src/sys

[Joerg Sonnenberger]

On Mon, Dec 04, 2017 at 10:14:33AM +0100, Manuel Bouyer wrote:
> On Mon, Dec 04, 2017 at 09:48:32AM +0100, Joerg Sonnenberger wrote:
> > On Fri, May 27, 2016 at 08:01:49PM +, Manuel Bouyer wrote:
> > > Module Name:  src
> > > Committed By: bouyer
> > > Date: Fri May 27 20:01:49 UTC 2016
> > > 
> > > Modified Files:
> > >   src/sys/arch/arm/allwinner: files.awin
> > >   src/sys/conf: files
> > >   src/sys/dev/ic: com.c comreg.h comvar.h ns16550reg.h
> > > 
> > > Log Message:
> > > The UART in the allwiner SoCs is not full-compatible with the 16550, and
> > > it's not a 16750 either. Like the 16750 it has the IIR_BUSY interrupt,
> > > which is triggered when writing to LCR while the chip
> > > can't accept it. But unlike the 16750, it has a specific register,
> > > HALT, to allow writing to the LCR and divisor registers, and then
> > > commit the changes.
> > > Tested on an A20 SoC, changing the baud rate while keeping the
> > > tty device open and incoming data.
> > 
> > Hello Manuel,
> > I'm not sure how you tested this, but you are trivally accessing an array
> > beyond the end. COM_REG_TFL and friends are all larger than the cr_map 
> > array.
> 
> Inded, but cr_map is used only if COM_REGMAP is defined, which is not
> the case for allwinner. That's why it works ...
> 
> I'll add a #error for this case.

That doesn't help since the code in com.c is compiled into the ARMADAXP
kernel.

Joerg

Re: CVS commit: src/sys

[Manuel Bouyer]

On Mon, Dec 04, 2017 at 09:48:32AM +0100, Joerg Sonnenberger wrote:
> On Fri, May 27, 2016 at 08:01:49PM +, Manuel Bouyer wrote:
> > Module Name:src
> > Committed By:   bouyer
> > Date:   Fri May 27 20:01:49 UTC 2016
> > 
> > Modified Files:
> > src/sys/arch/arm/allwinner: files.awin
> > src/sys/conf: files
> > src/sys/dev/ic: com.c comreg.h comvar.h ns16550reg.h
> > 
> > Log Message:
> > The UART in the allwiner SoCs is not full-compatible with the 16550, and
> > it's not a 16750 either. Like the 16750 it has the IIR_BUSY interrupt,
> > which is triggered when writing to LCR while the chip
> > can't accept it. But unlike the 16750, it has a specific register,
> > HALT, to allow writing to the LCR and divisor registers, and then
> > commit the changes.
> > Tested on an A20 SoC, changing the baud rate while keeping the
> > tty device open and incoming data.
> 
> Hello Manuel,
> I'm not sure how you tested this, but you are trivally accessing an array
> beyond the end. COM_REG_TFL and friends are all larger than the cr_map array.

Inded, but cr_map is used only if COM_REGMAP is defined, which is not
the case for allwinner. That's why it works ...

I'll add a #error for this case.

-- 
Manuel Bouyer 
 NetBSD: 26 ans d'experience feront toujours la difference
--

Re: CVS commit: src/sys

[Joerg Sonnenberger]

On Fri, May 27, 2016 at 08:01:49PM +, Manuel Bouyer wrote:
> Module Name:  src
> Committed By: bouyer
> Date: Fri May 27 20:01:49 UTC 2016
> 
> Modified Files:
>   src/sys/arch/arm/allwinner: files.awin
>   src/sys/conf: files
>   src/sys/dev/ic: com.c comreg.h comvar.h ns16550reg.h
> 
> Log Message:
> The UART in the allwiner SoCs is not full-compatible with the 16550, and
> it's not a 16750 either. Like the 16750 it has the IIR_BUSY interrupt,
> which is triggered when writing to LCR while the chip
> can't accept it. But unlike the 16750, it has a specific register,
> HALT, to allow writing to the LCR and divisor registers, and then
> commit the changes.
> Tested on an A20 SoC, changing the baud rate while keeping the
> tty device open and incoming data.

Hello Manuel,
I'm not sure how you tested this, but you are trivally accessing an array
beyond the end. COM_REG_TFL and friends are all larger than the cr_map array.

Joerg