OMAP 4430 SDP: rather sick with recent kernels

Wed Jan 14 15:10:08 PST 2015

On Wed, Jan 14, 2015 at 11:09:26AM -0800, Tony Lindgren wrote:
> * Russell King - ARM Linux <linux at arm.linux.org.uk> [150114 09:54]:
> > On Wed, Dec 17, 2014 at 09:52:52AM +0000, Russell King - ARM Linux wrote:
> > > The "combined" kernel boot follows a similar pattern, but yets a bit
> > > further before oopsing, with ASoC DAPM code printing random bits of
> > > kernel memory.
> > 
> > Note that the "combined" kernel (which is OMAP4 + Versatile Express)
> > has for a while now also spat out this:
> > 
> > WARNING: CPU: 0 PID: 1 at .../linux-build/drivers/bus/omap_l3_noc.c:147 l3_interrupt_handler+0x218/0x2f0()
> > 44000000.ocp:L3 Custom Error: MASTER MPU TARGET L4CFG (Idle): Data Access in Supervisor mode during Functional access
> > Modules linked in:
> > CPU: 0 PID: 1 Comm: swapper/0 Not tainted 3.19.0-rc4+ #1
> > Hardware name: Generic OMAP4 (Flattened Device Tree)
> > Backtrace:
> > [<c0011bdc>] (dump_backtrace) from [<c0011e9c>] (show_stack+0x18/0x1c)
> >  r6:c04c073e r5:00000009 r4:00000000 r3:00200040
> > [<c0011e84>] (show_stack) from [<c0466380>] (dump_stack+0x78/0x94)
> > [<c0466308>] (dump_stack) from [<c00379ec>] (warn_slowpath_common+0x8c/0xb8)
> >  r4:ce461b38 r3:ce458000
> > [<c0037960>] (warn_slowpath_common) from [<c0037abc>] (warn_slowpath_fmt+0x38/0x40)
> >  r8:c04c06a6 r7:c04c094c r6:80080003 r5:ce54d218 r4:f8000400
> > [<c0037a88>] (warn_slowpath_fmt) from [<c02098f8>] (l3_interrupt_handler+0x218/0x2f0)
> >  r3:ce54ad40 r2:c04c0772
> > [<c02096e0>] (l3_interrupt_handler) from [<c0076444>] (handle_irq_event_percpu+0x38/0x13c)
> >  r10:ce440700 r9:c06e1c58 r8:ce5b1960 r7:00000000 r6:00000000 r5:00000013
> >  r4:ce54f480
> > [<c007640c>] (handle_irq_event_percpu) from [<c007658c>] (handle_irq_event+0x44/0x64)
> >  r10:00000000 r9:ce5b1938 r8:ce5b1960 r7:00000001 r6:ce54f480 r5:ce440760
> >  r4:ce440700
> > [<c0076548>] (handle_irq_event) from [<c00795f4>] (handle_fasteoi_irq+0xb0/0x128)
> >  r6:ce440760 r5:c06c5fec r4:ce440700 r3:00000000
> > [<c0079544>] (handle_fasteoi_irq) from [<c0075d60>] (generic_handle_irq+0x28/0x38)
> >  r6:ce408000 r5:00000000 r4:00000013 r3:c0079544
> > [<c0075d38>] (generic_handle_irq) from [<c0075ebc>] (__handle_domain_irq+0x90/0xb8)
> >  r4:00000000 r3:00000110
> > [<c0075e2c>] (__handle_domain_irq) from [<c0008858>] (gic_handle_irq+0x44/0x68)
> >  r7:ce461d0c r6:c068b34c r5:ce461cd8 r4:fa240100
> > [<c0008814>] (gic_handle_irq) from [<c00129c4>] (__irq_svc+0x44/0x5c)
> > Exception stack(0xce461cd8 to 0xce461d20)
> > 1cc0:                                                       00000001 ce458508
> > 1ce0: 00000000 00000000 60000113 ce5b1960 0000002c 00000000 ce5b1960 ce5b1938
> > 1d00: 00000000 ce461d34 ce461cf0 ce461d20 c006d4f4 c046df64 20000113 ffffffff
> >  r6:ffffffff r5:20000113 r4:c046df64 r3:ce458000
> > [<c046df1c>] (_raw_spin_unlock_irqrestore) from [<c0077dd0>] (__setup_irq+0x3bc/0x4e4)
> >  r5:c06b7c00 r4:ce5b1900
> > [<c0077a14>] (__setup_irq) from [<c00780ec>] (setup_irq+0x50/0x90)
> >  r10:c06ea338 r9:20000113 r8:c06ea338 r7:ce67ec00 r6:c06b7c00 r5:0000002c
> >  r4:ce5b1900
> > [<c007809c>] (setup_irq) from [<c0032768>] (omap_system_dma_probe+0x20c/0x2a4)
> >  r6:ce67ec10 r5:0000002c r4:00000020 r3:00000002
> > [<c003255c>] (omap_system_dma_probe) from [<c0298788>] (platform_drv_probe+0x50/0x98)
> >  r10:c066d8b0 r9:00000000 r8:00000000 r7:c02971b8 r6:c06b7b94 r5:ffffffed
> >  r4:ce67ec10
> > [<c0298738>] (platform_drv_probe) from [<c0296fa8>] (driver_probe_device+0x13c/0x34c)
> >  r6:00000000 r5:c06b7b94 r4:ce67ec10 r3:c0298738
> > [<c0296e6c>] (driver_probe_device) from [<c0297230>] (__driver_attach+0x78/0x9c) r8:c0646228 r7:c02971b8 r6:c06b7b94 r5:ce67ec44 r4:ce67ec10
> > [<c02971b8>] (__driver_attach) from [<c0295424>] (bus_for_each_dev+0x5c/0x98)
> >  r6:c06b7b94 r5:ce461e48 r4:00000000 r3:c02971b8
> > [<c02953c8>] (bus_for_each_dev) from [<c02969c0>] (driver_attach+0x20/0x28)
> >  r7:00000000 r6:c06ce978 r5:ce61c100 r4:c06b7b94
> > [<c02969a0>] (driver_attach) from [<c02965ec>] (bus_add_driver+0xf8/0x1fc)
> > [<c02964f4>] (bus_add_driver) from [<c0297914>] (driver_register+0xa4/0xe8)
> >  r7:c06e9040 r6:c068d0e8 r5:c068d0e8 r4:c06b7b94
> > [<c0297870>] (driver_register) from [<c029861c>] (__platform_driver_register+0x50/0x64)
> >  r5:c068d0e8 r4:ce621f80
> > [<c02985cc>] (__platform_driver_register) from [<c0646240>] (omap_system_dma_init+0x18/0x20)
> > [<c0646228>] (omap_system_dma_init) from [<c0008c44>] (do_one_initcall+0x114/0x1e0)
> > [<c0008b30>] (do_one_initcall) from [<c0632e3c>] (kernel_init_freeable+0x110/0x1dc)
> >  r10:c066d8b0 r9:00000000 r8:0000007d r7:c06e9040 r6:c067ddf0 r5:c066d898
> >  r4:00000003
> > [<c0632d2c>] (kernel_init_freeable) from [<c0460840>] (kernel_init+0x10/0xec)
> >  r10:00000000 r8:00000000 r7:00000000 r6:00000000 r5:c0460830 r4:00000000
> > [<c0460830>] (kernel_init) from [<c000f0d0>] (ret_from_fork+0x14/0x24)
> >  r4:00000000 r3:00000000
> > ---[ end trace bcb85e0273c31888 ]---
> > OMAP DMA hardware revision 0.0
> > 
> > This does not occur when booting the plain "omap4" kernel.
> > 
> > I thought it may be related to another error which people have been
> > reporting, but as it's still there, I thought I should report it.
> > 
> > To me, this suggests that Versatile Express code may be initialising
> > on non-Versatile Express hardware, possibly touching hardware, but
> > it looks like it's happening within the setup_irq() called from the
> > legacy OMAP DMA code, though it's possible that could be because
> > Versatile Express code could be hitting some register which causes
> > OMAP4 to later to wrong.
> > 
> > I think it's going to be particularly horrid to debug...
> 
> This seems like it could be a similar issue to what we were seeing on
> omap3 where the legacy IRQ mappings went wrong without using the
> irq_domain_add_legacy(). Maybe vexpress affects NR_IRQS or something?
> 
> Also, I think we could now disable the legacy DMA init completely
> except for omap2 and 3.

I've added a "cat /proc/interrupts" which should reveal whether the IRQs
have changed between the two different builds.

-- 
FTTC broadband for 0.8mile line: currently at 10.5Mbps down 400kbps up
according to speedtest.net.