linux-stable.git/arch/m68k/atari, branch linux-4.9.y

m68k: Call timer_interrupt() with interrupts disabled

2020-01-29T09:24:36+00:00

[ Upstream commit 1efdd4bd254311498123a15fa0acd565f454da97 ]

Some platforms execute their timer handler with the interrupt priority
level set below 6. That means the handler could be interrupted by another
driver and this could lead to re-entry of the timer core.

Avoid this by use of local_irq_save/restore for timer interrupt dispatch.
This provides mutual exclusion around the timer interrupt flag access
which is needed later in this series for the clocksource conversion.

Reported-by: Thomas Gleixner 
Link: http://lkml.kernel.org/r/alpine.DEB.2.21.1811131407120.2697@nanos.tec.linutronix.de
Signed-off-by: Finn Thain 
Signed-off-by: Geert Uytterhoeven 
Signed-off-by: Sasha Levin

m68k: Use IS_ENABLED() instead of checking for built-in or module

2016-09-19T09:29:45+00:00

Signed-off-by: Geert Uytterhoeven

m68k/atari, m68k/sun3: Fix SCSI platform device registration when driver is modular

2016-01-10T12:56:40+00:00

Fixes: 3ff228af84b5 ("atari_scsi: Convert to platform device")
Fixes: 0d31f8759109 ("sun3_scsi: Convert to platform device")
Reported-by: Michael Schmitz 
Signed-off-by: Finn Thain 
Signed-off-by: Geert Uytterhoeven

m68k/atari: Remove obsolete IRQ_TYPE_*

2015-01-15T12:44:51+00:00

IRQ_TYPE_SLOW, IRQ_TYPE_FAST, and IRQ_TYPE_PRIO are no longer used by
the Atari platform interrupt code since commit 734085651c9b80aa
("[PATCH] m68k: convert atari irq code") in v2.6.18-rc1, so drop them.

Note that their values have been reused for different purposes
(IRQ_TYPE_NONE, IRQ_TYPE_EDGE_RISING, and IRQ_TYPE_EDGE_FALLING) since
commit 6a6de9ef5850d063 ("[PATCH] genirq: core") in v2.6.18-rc1.

Signed-off-by: Geert Uytterhoeven

m68k/atari: Remove obsolete keyboard_tasklet scheduling

2015-01-13T08:44:17+00:00

If CONFIG_VT=n:

arch/m68k/atari/built-in.o: In function `atari_keyboard_interrupt':
atakeyb.c:(.text+0x1846): undefined reference to `keyboard_tasklet'
atakeyb.c:(.text+0x1852): undefined reference to `keyboard_tasklet'

I think the keyboard_tasklet scheduling is no longer needed, as I
believe it's handled by drivers/tty/vt/keyboard.c based on events
received from the input subsystem. So just remove it.

Signed-off-by: Geert Uytterhoeven 
Tested-by: Michael Schmitz

m68k/atari: atakeyb.c - Remove some unused functions

2015-01-11T10:38:44+00:00

Remove some functions that are not used anywhere:
atari_kbd_leds() ikbd_exec() ikbd_mem_read() ikbd_mem_write()
ikbd_clock_get() ikbd_clock_set() ikbd_pause() ikbd_resume()

This was partially found by using a static code analysis program called
cppcheck.

Signed-off-by: Rickard Strandqvist 
Signed-off-by: Geert Uytterhoeven

atari_scsi: Convert to platform device

2014-11-20T08:11:14+00:00

Convert atari_scsi to platform device and eliminate scsi_register().

Validate __setup options later on so that module options are checked as well.

Remove the comment about the scsi mid-layer disabling the host irq as it
is no longer true (AFAICT). Also remove the obsolete slow interrupt stuff
(IRQ_TYPE_SLOW == 0 anyway).

Signed-off-by: Finn Thain 
Reviewed-by: Hannes Reinecke 
Tested-by: Michael Schmitz 
Acked-by: Geert Uytterhoeven 
Signed-off-by: Christoph Hellwig

atari_scsi: Fix atari_scsi deadlocks on Falcon

2014-11-20T08:11:13+00:00

Don't disable irqs when waiting for the ST DMA "lock"; its release may
require an interrupt.

Introduce stdma_try_lock() for use in soft irq context. atari_scsi now tells
the SCSI mid-layer to defer queueing a command if the ST DMA lock is not
available, as per Michael's patch:
http://marc.info/?l=linux-m68k&m=139095335824863&w=2

The falcon_got_lock variable is race prone: we can't disable IRQs while
waiting to acquire the lock, so after acquiring it there must be some
interval during which falcon_got_lock remains false. Introduce
stdma_is_locked_by() to replace falcon_got_lock.

The falcon_got_lock tests in the EH handlers are incorrect these days. It
can happen that an EH handler is called after a command completes normally.
Remove these checks along with falcon_got_lock.

Also remove the complicated and racy fairness wait queues. If fairness is an
issue (when SCSI competes with IDE for the ST DMA interrupt), the solution
is likely to be a lower value for host->can_queue.

Signed-off-by: Finn Thain 
Reviewed-by: Hannes Reinecke 
Tested-by: Michael Schmitz 
Acked-by: Geert Uytterhoeven 
Signed-off-by: Christoph Hellwig

m68k/atari - stram: Add missing #include

2014-09-14T09:51:35+00:00

If CONFIG_BLOCK=n:

arch/m68k/atari/stram.c:44: error: variable ‘stram_pool’ has initializer but incomplete type
arch/m68k/atari/stram.c:45: error: unknown field ‘name’ specified in initializer
arch/m68k/atari/stram.c:46: warning: excess elements in struct initializer
arch/m68k/atari/stram.c:46: warning: (near initialization for ‘stram_pool’)
arch/m68k/atari/stram.c: In function ‘atari_stram_reserve_pages’:
arch/m68k/atari/stram.c:97: error: invalid use of undefined type ‘struct resource’
...

Signed-off-by: Geert Uytterhoeven

m68k/atari - stram: alloc ST-RAM pool even if kernel not in ST-RAM

2014-05-26T20:41:24+00:00

With the kernel loaded to FastRAM (TT-RAM), none of the ST-RAM
address range is mapped by init_mem, and ST-RAM is not accessible
through the normal allocation pathways as a result.

Implement ST-RAM pool allocation to be based on physical addresses
always (it already was when the kernel was loaded in ST-RAM).
Return kernel virtual addresses as per normal.

The current test for the kernel residing in ST-RAM always returns
true. Use the bootinfo memory chunk order instead - with the kernel
in FastRAM, ST-RAM (phys. 0x0) is not the first chunk.

In case the kernel is running from FastRAM, delay mapping of ST-RAM
pool until after mem_init.

Provide helper functions for those users of ST-RAM that need
to be aware of the backing physical addresses.

Kudos to Geert for his hints on getting this started.

Signed-off-by: Michael Schmitz 
Signed-off-by: Geert Uytterhoeven