mbox series

[RFC,0/2] vlynq: remove bus support

Message ID 20230916091125.3221-1-wsa@kernel.org (mailing list archive)
Headers show
Series vlynq: remove bus support | expand

Message

Wolfram Sang Sept. 16, 2023, 9:11 a.m. UTC
From: Wolfram Sang <wsa+renesas@sang-engineering.com>

I'd like to clean up the 'drivers' directory and Kconfig menu, at least
a little. One major thing I noticed is that VLYNQ can actually be
removed. From patch 2:

---
There are no users with a vlynq_driver in the Kernel tree. So, remove
the bus driver which hardly has seen any activity since 2009. It was
even marked EXPERIMENTAL as long as that symbol existed. OpenWRT had
some out-of-tree drivers which they probably intended to upport, but AR7
devices are not supported anymore because they are "stuck with Kernel
3.18" [1]. So, this code can go nowadays.

[1] https://openwrt.org/docs/techref/targets/ar7
---

Patch 1 removes MIPS specific bus initialization code which is the only
user of the VLYNQ bus currently.

Sending out as RFC to gather comments. Patches are based on 6.6-rc1 and
created with "--irreversible-delete" to save some space. They are
compile tested only, buildbots were happy.

I think this mainly goes to Florian Fainelli. Florian, what do you
think?

All the best!

   Wolfram


Wolfram Sang (2):
  MIPS: AR7: remove VLYNQ init
  vlynq: remove bus driver

 MAINTAINERS                          |   7 -
 arch/mips/Kconfig                    |   1 -
 arch/mips/ar7/platform.c             | 193 +------
 arch/mips/include/asm/mach-ar7/ar7.h |   1 -
 drivers/Kconfig                      |   2 -
 drivers/Makefile                     |   1 -
 drivers/vlynq/Kconfig                |  21 -
 drivers/vlynq/Makefile               |   6 -
 drivers/vlynq/vlynq.c                | 799 ---------------------------
 include/linux/vlynq.h                | 149 -----
 10 files changed, 1 insertion(+), 1179 deletions(-)
 delete mode 100644 drivers/vlynq/Kconfig
 delete mode 100644 drivers/vlynq/Makefile
 delete mode 100644 drivers/vlynq/vlynq.c
 delete mode 100644 include/linux/vlynq.h

Comments

Jonas Gorski Sept. 16, 2023, 11:11 a.m. UTC | #1
On Sat, 16 Sept 2023 at 11:18, Wolfram Sang <wsa@kernel.org> wrote:
>
> From: Wolfram Sang <wsa+renesas@sang-engineering.com>
>
> I'd like to clean up the 'drivers' directory and Kconfig menu, at least
> a little. One major thing I noticed is that VLYNQ can actually be
> removed. From patch 2:
>
> ---
> There are no users with a vlynq_driver in the Kernel tree. So, remove
> the bus driver which hardly has seen any activity since 2009. It was
> even marked EXPERIMENTAL as long as that symbol existed. OpenWRT had
> some out-of-tree drivers which they probably intended to upport, but AR7
> devices are not supported anymore because they are "stuck with Kernel
> 3.18" [1]. So, this code can go nowadays.
>
> [1] https://openwrt.org/docs/techref/targets/ar7
> ---
>
> Patch 1 removes MIPS specific bus initialization code which is the only
> user of the VLYNQ bus currently.
>
> Sending out as RFC to gather comments. Patches are based on 6.6-rc1 and
> created with "--irreversible-delete" to save some space. They are
> compile tested only, buildbots were happy.
>
> I think this mainly goes to Florian Fainelli. Florian, what do you
> think?

I'm not Florian, but I worked a bit with AR7 as well, and going
through the (previously) supported/known devices they topped out at 8
MiB Flash and 32 MiB RAM, which isn't enough to run any reasonably
modern OpenWrt even if one would be willing to re-add support for it.

So from my perspective I suggest even dropping the AR7 code entirely.
I seriously doubt anyone is running a modern kernel on that.

Best Regards,
Jonas
Greg Kroah-Hartman Sept. 16, 2023, 11:43 a.m. UTC | #2
On Sat, Sep 16, 2023 at 11:11:22AM +0200, Wolfram Sang wrote:
> From: Wolfram Sang <wsa+renesas@sang-engineering.com>
> 
> I'd like to clean up the 'drivers' directory and Kconfig menu, at least
> a little. One major thing I noticed is that VLYNQ can actually be
> removed. From patch 2:
> 
> ---
> There are no users with a vlynq_driver in the Kernel tree. So, remove
> the bus driver which hardly has seen any activity since 2009. It was
> even marked EXPERIMENTAL as long as that symbol existed. OpenWRT had
> some out-of-tree drivers which they probably intended to upport, but AR7
> devices are not supported anymore because they are "stuck with Kernel
> 3.18" [1]. So, this code can go nowadays.
> 
> [1] https://openwrt.org/docs/techref/targets/ar7
> ---
> 
> Patch 1 removes MIPS specific bus initialization code which is the only
> user of the VLYNQ bus currently.
> 
> Sending out as RFC to gather comments. Patches are based on 6.6-rc1 and
> created with "--irreversible-delete" to save some space. They are
> compile tested only, buildbots were happy.
> 
> I think this mainly goes to Florian Fainelli. Florian, what do you
> think?

Acked-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Florian Fainelli Sept. 16, 2023, 4:42 p.m. UTC | #3
On 9/16/2023 4:11 AM, Jonas Gorski wrote:
> On Sat, 16 Sept 2023 at 11:18, Wolfram Sang <wsa@kernel.org> wrote:
>>
>> From: Wolfram Sang <wsa+renesas@sang-engineering.com>
>>
>> I'd like to clean up the 'drivers' directory and Kconfig menu, at least
>> a little. One major thing I noticed is that VLYNQ can actually be
>> removed. From patch 2:
>>
>> ---
>> There are no users with a vlynq_driver in the Kernel tree. So, remove
>> the bus driver which hardly has seen any activity since 2009. It was
>> even marked EXPERIMENTAL as long as that symbol existed. OpenWRT had
>> some out-of-tree drivers which they probably intended to upport, but AR7
>> devices are not supported anymore because they are "stuck with Kernel
>> 3.18" [1]. So, this code can go nowadays.
>>
>> [1] https://openwrt.org/docs/techref/targets/ar7
>> ---
>>
>> Patch 1 removes MIPS specific bus initialization code which is the only
>> user of the VLYNQ bus currently.
>>
>> Sending out as RFC to gather comments. Patches are based on 6.6-rc1 and
>> created with "--irreversible-delete" to save some space. They are
>> compile tested only, buildbots were happy.
>>
>> I think this mainly goes to Florian Fainelli. Florian, what do you
>> think?
> 
> I'm not Florian, but I worked a bit with AR7 as well, and going
> through the (previously) supported/known devices they topped out at 8
> MiB Flash and 32 MiB RAM, which isn't enough to run any reasonably
> modern OpenWrt even if one would be willing to re-add support for it.
> 
> So from my perspective I suggest even dropping the AR7 code entirely.
> I seriously doubt anyone is running a modern kernel on that.

Agreed, TI AR7 is nearly 25 years old now, we should be able to remove 
that. Wolfram, do you feel like doing that or would you rather have me 
do it, say next week?
Wolfram Sang Sept. 16, 2023, 5:02 p.m. UTC | #4
> Agreed, TI AR7 is nearly 25 years old now, we should be able to remove that.
> Wolfram, do you feel like doing that or would you rather have me do it, say
> next week?

I will try it, thanks for the heads up everyone!
Wolfram Sang Sept. 18, 2023, 1:18 p.m. UTC | #5
Hi all,

> Agreed, TI AR7 is nearly 25 years old now, we should be able to remove that.
> Wolfram, do you feel like doing that or would you rather have me do it, say
> next week?

So, I pushed out an RC here:

git://git.kernel.org/pub/scm/linux/kernel/git/wsa/linux.git renesas/cleanup/remove_ar7

Just waiting for the buildbot report before I send out patches. buildbot
was already happy with my protoype, though. But I needed to reorder the
patches. (I do wonder, though, if they build AR7. I would have expected
a build failure with my prototype).

Happy hacking,

   Wolfram
Florian Fainelli Sept. 18, 2023, 6:33 p.m. UTC | #6
On 9/18/23 06:18, Wolfram Sang wrote:
> Hi all,
> 
>> Agreed, TI AR7 is nearly 25 years old now, we should be able to remove that.
>> Wolfram, do you feel like doing that or would you rather have me do it, say
>> next week?
> 
> So, I pushed out an RC here:
> 
> git://git.kernel.org/pub/scm/linux/kernel/git/wsa/linux.git renesas/cleanup/remove_ar7
> 
> Just waiting for the buildbot report before I send out patches. buildbot
> was already happy with my protoype, though. But I needed to reorder the
> patches. (I do wonder, though, if they build AR7. I would have expected
> a build failure with my prototype).

Looks good to me, just a few nits:

arch/arm/configs/pxa_defconfig:CONFIG_MTD_AR7_PARTS=m

this is clearly a stray reference to begin with, but should be folded in 
your commit removing the MTD partition parser.

Other than that, LGTM and thanks for doing that clean up.
Wolfram Sang Sept. 19, 2023, 6:37 a.m. UTC | #7
Hi Florian,

> arch/arm/configs/pxa_defconfig:CONFIG_MTD_AR7_PARTS=m

Thanks, I removed this line as well.

> Other than that, LGTM and thanks for doing that clean up.

Cool, thanks for the review!

Happy hacking,

   Wolfram