Re: Please Add Marvell mv78xx0 flavour

2012-02-04 Thread Steve McIntyre
On Thu, Feb 02, 2012 at 03:34:35AM +, Ben Hutchings wrote:
On Wed, 2012-02-01 at 14:55 +, Steve McIntyre wrote:
 On Thu, Jan 26, 2012 at 04:20:20AM +, Steve McIntyre wrote:
 On Thu, Jan 26, 2012 at 03:36:53AM +, Ben Hutchings wrote:
 On Wed, 2012-01-25 at 15:49 +, Steve McIntyre wrote:
  Following up on the conversation from last week...
  
  Attached is a diff for squeeze. This kernel boots and runs fine on
  arne.debian.org, just tested. The diff between the new config.mv78xx0
  and config.orion5x is tiny (mostly comments), but they're not
  compatible as a single flavour:
 
 Right.  But I look at these two configurations and cannot understand the
 reasoning behind the driver selection.
 
 They have PCI, but a lot of PCI drivers disabled.  I assume this is
 because there are no PCI slots?  If that is the case then there are a
 whole lot more PCI drivers that cannot be used and should be disabled,
 which will cut the compile time down significantly.
 
 Ah, no. They have PCI slots and could probably support most PCI
 things, I just didn't consider that so far, basing on config.orion5x.
 Is it preferred to add everything that might work for these boards, or
 acceptable to go for a more minimal config to help reduce build time?
 
 Ping?

I will apply your configuration to unstable now, but I think we need to
review which drivers should be enabled.  I realise you're trying to
match the existing 'orion5x' but let's not compound the mistake.

OK, no problem. Let's talk about that at some point soon. Or... is it
just as simple as adding all the PCI-based options?

If you still want this flavour to be added to stable as well then, as
with any stable update, you need to open a bug report.

Fine.

Your descriptions weren't consistent.  The flavour is 'mv78xx0' but you
used 'DB-78x00-DP' (1 x) in the long description.  Then the short
description said 'Feroceon', but that's just the CPU core and it is also
used in the kirkwood and orion5x flavours.  So I changed those all to
'78xx0'.

Cool, thanks!

-- 
Steve McIntyre, Cambridge, UK.st...@einval.com
  Mature Sporty Personal
  More Innovation More Adult
  A Man in Dandism
  Powered Midship Specialty


-- 
To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/20120203135858.gj5...@einval.com



Re: Please Add Marvell mv78xx0 flavour

2012-02-01 Thread Steve McIntyre
On Thu, Jan 26, 2012 at 04:20:20AM +, Steve McIntyre wrote:
On Thu, Jan 26, 2012 at 03:36:53AM +, Ben Hutchings wrote:
On Wed, 2012-01-25 at 15:49 +, Steve McIntyre wrote:
 Following up on the conversation from last week...
 
 Attached is a diff for squeeze. This kernel boots and runs fine on
 arne.debian.org, just tested. The diff between the new config.mv78xx0
 and config.orion5x is tiny (mostly comments), but they're not
 compatible as a single flavour:

Right.  But I look at these two configurations and cannot understand the
reasoning behind the driver selection.

They have PCI, but a lot of PCI drivers disabled.  I assume this is
because there are no PCI slots?  If that is the case then there are a
whole lot more PCI drivers that cannot be used and should be disabled,
which will cut the compile time down significantly.

Ah, no. They have PCI slots and could probably support most PCI
things, I just didn't consider that so far, basing on config.orion5x.
Is it preferred to add everything that might work for these boards, or
acceptable to go for a more minimal config to help reduce build time?

Ping?

-- 
Steve McIntyre, Cambridge, UK.st...@einval.com
C++ ate my sanity -- Jon Rabone


-- 
To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/20120201145531.gc7...@einval.com



Re: Please Add Marvell mv78xx0 flavour

2012-02-01 Thread Ben Hutchings
On Wed, 2012-02-01 at 14:55 +, Steve McIntyre wrote:
 On Thu, Jan 26, 2012 at 04:20:20AM +, Steve McIntyre wrote:
 On Thu, Jan 26, 2012 at 03:36:53AM +, Ben Hutchings wrote:
 On Wed, 2012-01-25 at 15:49 +, Steve McIntyre wrote:
  Following up on the conversation from last week...
  
  Attached is a diff for squeeze. This kernel boots and runs fine on
  arne.debian.org, just tested. The diff between the new config.mv78xx0
  and config.orion5x is tiny (mostly comments), but they're not
  compatible as a single flavour:
 
 Right.  But I look at these two configurations and cannot understand the
 reasoning behind the driver selection.
 
 They have PCI, but a lot of PCI drivers disabled.  I assume this is
 because there are no PCI slots?  If that is the case then there are a
 whole lot more PCI drivers that cannot be used and should be disabled,
 which will cut the compile time down significantly.
 
 Ah, no. They have PCI slots and could probably support most PCI
 things, I just didn't consider that so far, basing on config.orion5x.
 Is it preferred to add everything that might work for these boards, or
 acceptable to go for a more minimal config to help reduce build time?
 
 Ping?

I will apply your configuration to unstable now, but I think we need to
review which drivers should be enabled.  I realise you're trying to
match the existing 'orion5x' but let's not compound the mistake.

If you still want this flavour to be added to stable as well then, as
with any stable update, you need to open a bug report.

Your descriptions weren't consistent.  The flavour is 'mv78xx0' but you
used 'DB-78x00-DP' (1 x) in the long description.  Then the short
description said 'Feroceon', but that's just the CPU core and it is also
used in the kirkwood and orion5x flavours.  So I changed those all to
'78xx0'.

Ben.

-- 
Ben Hutchings
Lowery's Law:
 If it jams, force it. If it breaks, it needed replacing anyway.


signature.asc
Description: This is a digitally signed message part


Re: Please Add Marvell mv78xx0 flavour

2012-01-26 Thread Steve McIntyre
On Thu, Jan 26, 2012 at 04:20:20AM +, Steve McIntyre wrote:
On Thu, Jan 26, 2012 at 03:36:53AM +, Ben Hutchings wrote:
On Wed, 2012-01-25 at 15:49 +, Steve McIntyre wrote:
 Following up on the conversation from last week...
 
 Attached is a diff for squeeze. This kernel boots and runs fine on
 arne.debian.org, just tested. The diff between the new config.mv78xx0
 and config.orion5x is tiny (mostly comments), but they're not
 compatible as a single flavour:

Right.  But I look at these two configurations and cannot understand the
reasoning behind the driver selection.

They have PCI, but a lot of PCI drivers disabled.  I assume this is
because there are no PCI slots?  If that is the case then there are a
whole lot more PCI drivers that cannot be used and should be disabled,
which will cut the compile time down significantly.

Ah, no. They have PCI slots and could probably support most PCI
things, I just didn't consider that so far, basing on config.orion5x.
Is it preferred to add everything that might work for these boards, or
acceptable to go for a more minimal config to help reduce build time?

In the meantime, a very similar set of changes for current sid/trunk
also works for the mv78xx0 boards (see attached). The diff from the
orion5x config is even smaller:

e102122-lin:~/debian/marvell/debian-sid/config/armel$ diff -u config.orion5x 
config.mv78xx0 | wc -l
68

-- 
Steve McIntyre, Cambridge, UK.st...@einval.com
  Mature Sporty Personal
  More Innovation More Adult
  A Man in Dandism
  Powered Midship Specialty


sid-mv78xx0.gz
Description: Binary data


Please Add Marvell mv78xx0 flavour

2012-01-25 Thread Steve McIntyre
Following up on the conversation from last week...

Attached is a diff for squeeze. This kernel boots and runs fine on
arne.debian.org, just tested. The diff between the new config.mv78xx0
and config.orion5x is tiny (mostly comments), but they're not
compatible as a single flavour:

$ diff -u config.orion5x config.mv78xx0 | wc -l
112

Looking into an equivalent for sid right now.

-- 
Steve McIntyre, Cambridge, UK.st...@einval.com
I've only once written 'SQL is my bitch' in a comment. But that code 
 is in use on a military site... -- Simon Booth


squeeze-mv78xx0.gz
Description: Binary data


Re: Please Add Marvell mv78xx0 flavour

2012-01-25 Thread Steve McIntyre
In article 20120125154946.ga21...@einval.com you write:
-=-=-=-=-=-

Following up on the conversation from last week...

Attached is a diff for squeeze. This kernel boots and runs fine on
arne.debian.org, just tested. The diff between the new config.mv78xx0
and config.orion5x is tiny (mostly comments), but they're not
compatible as a single flavour:

$ diff -u config.orion5x config.mv78xx0 | wc -l
112

Looking into an equivalent for sid right now.

Gah, and forgot the following to turn on HIGHMEM for the Marvell
machines.

93sam@abel:~/debian/config/armel$ diff -u config.mv78xx0~ config.mv78xx0
--- config.mv78xx0~ 2012-01-22 17:45:31.0 +
+++ config.mv78xx0  2012-01-25 16:24:40.0 +
@@ -79,6 +79,7 @@
 # CONFIG_CPU_DCACHE_DISABLE is not set
 # CONFIG_CPU_DCACHE_WRITETHROUGH is not set
 CONFIG_CACHE_FEROCEON_L2=y
+CONFIG_HIGHMEM=y
 
 ##
 ## file: drivers/ata/Kconfig


-- 
Steve McIntyre, Cambridge, UK.st...@einval.com
Because heaters aren't purple! -- Catherine Pitt


-- 
To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/e1rqcze-0008ew...@mail.einval.com



Re: Please Add Marvell mv78xx0 flavour

2012-01-25 Thread Ben Hutchings
On Wed, 2012-01-25 at 15:49 +, Steve McIntyre wrote:
 Following up on the conversation from last week...
 
 Attached is a diff for squeeze. This kernel boots and runs fine on
 arne.debian.org, just tested. The diff between the new config.mv78xx0
 and config.orion5x is tiny (mostly comments), but they're not
 compatible as a single flavour:

Right.  But I look at these two configurations and cannot understand the
reasoning behind the driver selection.

They have PCI, but a lot of PCI drivers disabled.  I assume this is
because there are no PCI slots?  If that is the case then there are a
whole lot more PCI drivers that cannot be used and should be disabled,
which will cut the compile time down significantly.

Ben.

 $ diff -u config.orion5x config.mv78xx0 | wc -l
 112
 
 Looking into an equivalent for sid right now.
 

-- 
Ben Hutchings
Horngren's Observation:
   Among economists, the real world is often a special case.


signature.asc
Description: This is a digitally signed message part


Re: Please Add Marvell mv78xx0 flavour

2012-01-25 Thread Steve McIntyre
On Thu, Jan 26, 2012 at 03:36:53AM +, Ben Hutchings wrote:
On Wed, 2012-01-25 at 15:49 +, Steve McIntyre wrote:
 Following up on the conversation from last week...
 
 Attached is a diff for squeeze. This kernel boots and runs fine on
 arne.debian.org, just tested. The diff between the new config.mv78xx0
 and config.orion5x is tiny (mostly comments), but they're not
 compatible as a single flavour:

Right.  But I look at these two configurations and cannot understand the
reasoning behind the driver selection.

They have PCI, but a lot of PCI drivers disabled.  I assume this is
because there are no PCI slots?  If that is the case then there are a
whole lot more PCI drivers that cannot be used and should be disabled,
which will cut the compile time down significantly.

Ah, no. They have PCI slots and could probably support most PCI
things, I just didn't consider that so far, basing on config.orion5x.
Is it preferred to add everything that might work for these boards, or
acceptable to go for a more minimal config to help reduce build time?

-- 
Steve McIntyre, Cambridge, UK.st...@einval.com
I can't ever sleep on planes ... call it irrational if you like, but I'm
 afraid I'll miss my stop -- Vivek Dasmohapatra


-- 
To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/20120126042020.gd5...@einval.com