Re: [PATCH] staging: tidspbridge: enable watchdog by default

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On Wed, Feb 1, 2012 at 5:44 AM, Greg KH <greg@xxxxxxxxx> wrote:
> On Wed, Feb 01, 2012 at 03:37:55AM +0200, Felipe Contreras wrote:
>> On Wed, Feb 1, 2012 at 3:12 AM, Greg KH <gregkh@xxxxxxx> wrote:
>> > On Wed, Feb 01, 2012 at 03:01:32AM +0200, Felipe Contreras wrote:
>> >> From: Felipe Contreras <felipe.contreras@xxxxxxxxx>
>> >>
>> >> The public images have it enabled, it's safer, and we get rid of this
>> >> warning:
>> >>
>> >> WARNING: at arch/arm/mach-omap2/omap_l3_smx.c:162 omap3_l3_app_irq+0x114/0x15c()
>> >> In-band Error seen by IVA_SS  at address 0
>> >> Modules linked in: bridgedriver(C+) mailbox_mach mailbox ipv6 g_ether
>> >> [<c0012848>] (unwind_backtrace+0x0/0xec) from [<c002e760>] (warn_slowpath_common+0x4c/0x64)
>> >> [<c002e760>] (warn_slowpath_common+0x4c/0x64) from [<c002e7f8>] (warn_slowpath_fmt+0x2c/0x3c)
>> >> [<c002e7f8>] (warn_slowpath_fmt+0x2c/0x3c) from [<c00217c8>] (omap3_l3_app_irq+0x114/0x15c)
>> >> [<c00217c8>] (omap3_l3_app_irq+0x114/0x15c) from [<c0059844>] (handle_irq_event_percpu+0x28/0x174)
>> >> [<c0059844>] (handle_irq_event_percpu+0x28/0x174) from [<c00599b8>] (handle_irq_event+0x28/0x38)
>> >> [<c00599b8>] (handle_irq_event+0x28/0x38) from [<c005b8d0>] (handle_level_irq+0xb8/0xe0)
>> >> [<c005b8d0>] (handle_level_irq+0xb8/0xe0) from [<c0059598>] (generic_handle_irq+0x28/0x30)
>> >> [<c0059598>] (generic_handle_irq+0x28/0x30) from [<c000e62c>] (handle_IRQ+0x60/0x84)
>> >> [<c000e62c>] (handle_IRQ+0x60/0x84) from [<c000d334>] (__irq_svc+0x34/0x80)
>> >> [<c000d334>] (__irq_svc+0x34/0x80) from [<c0015804>] (v7_dma_inv_range+0x34/0x48)
>> >> [<c0015804>] (v7_dma_inv_range+0x34/0x48) from [<c00129ec>] (dma_cache_maint_page+0x2c/0x34)
>> >> [<c00129ec>] (dma_cache_maint_page+0x2c/0x34) from [<c0012a34>] (___dma_page_dev_to_cpu+0x24/0x68)
>> >> [<c0012a34>] (___dma_page_dev_to_cpu+0x24/0x68) from [<c0012b50>] (dma_unmap_sg+0x40/0x64)
>> >> [<c0012b50>] (dma_unmap_sg+0x40/0x64) from [<c019fba4>] (omap_hsmmc_post_req+0x3c/0x48)
>> >> [<c019fba4>] (omap_hsmmc_post_req+0x3c/0x48) from [<c01932f4>] (mmc_post_req+0x18/0x1c)
>> >> [<c01932f4>] (mmc_post_req+0x18/0x1c) from [<c019535c>] (mmc_start_req+0xd4/0xec)
>> >> [<c019535c>] (mmc_start_req+0xd4/0xec) from [<c019d330>] (mmc_blk_issue_rw_rq+0x64/0x244)
>> >> [<c019d330>] (mmc_blk_issue_rw_rq+0x64/0x244) from [<c019d7bc>] (mmc_blk_issue_rq+0x2ac/0x2cc)
>> >> [<c019d7bc>] (mmc_blk_issue_rq+0x2ac/0x2cc) from [<c019e1c0>] (mmc_queue_thread+0x8c/0xf8)
>> >> [<c019e1c0>] (mmc_queue_thread+0x8c/0xf8) from [<c0044c10>] (kthread+0x84/0x8c)
>> >> [<c0044c10>] (kthread+0x84/0x8c) from [<c000e694>] (kernel_thread_exit+0x0/0x8)
>> >> ---[ end trace 5dec1c8d7857375d ]---
>> >>
>> >> Signed-off-by: Felipe Contreras <felipe.contreras@xxxxxxxxx>
>> >> ---
>> >>  drivers/staging/tidspbridge/Kconfig |    1 +
>> >>  1 files changed, 1 insertions(+), 0 deletions(-)
>> >>
>> >> diff --git a/drivers/staging/tidspbridge/Kconfig b/drivers/staging/tidspbridge/Kconfig
>> >> index 21a559e..614e974 100644
>> >> --- a/drivers/staging/tidspbridge/Kconfig
>> >> +++ b/drivers/staging/tidspbridge/Kconfig
>> >> @@ -61,6 +61,7 @@ config TIDSPBRIDGE_CACHE_LINE_CHECK
>> >>  config TIDSPBRIDGE_WDT3
>> >>       bool "Enable watchdog timer"
>> >>       depends on TIDSPBRIDGE
>> >> +     default y
>> >
>> > Nothing is ever default y, sorry, I can't accept this.
>>
>> % git grep 'default y' -- 'drivers/*/Kconfig' | wc -l
>> 493
>
> That's nice, but proves nothing, no new entries should be y.
>
> If it should be y, then it should just not be an option and always
> enabled, right?  Why not do that?

What if I compiled a DSP baseimage that doesn't have this watchdog?

And using that logic, we should start removing all configurations in
the kernel that have 'default y'.

No. There's a reason they are *configurations*, and there's a reason
they are 'default y'.

You might have noticed Linus' rant about ARM defconfigs (LWN
article[1]), and the solution to that was simplified configs (make
savedefconfig), and the ideal would be to have no defconfigs at all.
But in order to achieve that ideal goal (or at least move in that
direction), the Kconfig files need to have good defaults--this was
also agreed in the discussion.

If you look at arch/arm/configs/, there are still quite a lot of
configs there, and many not small at all. If Linus goes forward with
his threat to remove ARM defconfigs, we would be in a bit of a
situation, specially since there's still a lot of Kconfigs that have
not been fixed.

There's plenty of other reasons why 'default y' is good.

This change is good, it helps everyone, and it doesn't hurt anyone.
But if you like, I can put forward a case and take it with the penguin
chief.

[1] http://lwn.net/Articles/391372/

-- 
Felipe Contreras
--
To unsubscribe from this list: send the line "unsubscribe linux-omap" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html


[Index of Archives]     [Linux Arm (vger)]     [ARM Kernel]     [ARM MSM]     [Linux Tegra]     [Linux WPAN Networking]     [Linux Wireless Networking]     [Maemo Users]     [Linux USB Devel]     [Video for Linux]     [Linux Audio Users]     [Yosemite Trails]     [Linux Kernel]     [Linux SCSI]

  Powered by Linux