Fengguang Wu reports that building ARM with !MMU results in the
following build error:
arch/arm/kernel/built-in.o: In function `__soft_restart':
>> :(.text+0x1624): undefined reference to `arch_virt_to_idmap'
Fix this by adding an appropriate IS_ENABLED(CONFIG_MMU) into the
__virt_to_idmap() inline function.
Signed-off-by: Russell King <[email protected]>
We must invalidate the L1 cache before enabling coherency, otherwise
secondary CPUs can inject invalid cache lines into the coherent CPU
cluster, which could then be migrated to other CPUs. This fixes a
recent regression with SoCFPGA randomly failing to boot.
Fixes: 02b4e2756e ("ARM: v7 setup function should invalidate L1 cache")
Signed-off-by: Russell King <[email protected]>
We're not supposed to store the adjusted mode into crtc->mode. We don't
use it anyway, so we can safely remove this.
Signed-off-by: Russell King <[email protected]>
Fix the start address calculation when overlay is partially off screen.
fb->bits_per_pixel is not set for YUV formats, and is always zero, which
led to the first component always starting at zero.
Use drm_format_plane_cpp() instead.
This also revealed a problem in that YUYV formats toggle the U/V data
for odd pixel start address offsets. We try to rectify that by
toggling the U/V swap, which for the most part works, but seemingly
introduces a flicker for one scan frame of swapped U/V.
However, these changes result in an overall improvement.
Signed-off-by: Russell King <[email protected]>
Use drm_plane_helper_check_update() rather than our own code to validate
and limit the size of the displayed image. As we are able to support
scaling, permit the full scaling ability.
Signed-off-by: Russell King <[email protected]>
Fix the gem object freeing after a partial import of a dma buffer,
eg, one which has been imported, but not mapped. This was provoking
a warning from the dma_buf code.
Signed-off-by: Russell King <[email protected]>
The Armada overlay plane wasn't being properly cleaned up as it was
missing a call to drm_plane_cleanup(). It also wasn't freeing the
right type of pointer (although we were still freeing the right
pointer value.)
Signed-off-by: Russell King <[email protected]>
Nothing was waking up the overlay plane wait queue, so we were fully
reliant on the HZ/25 wait timing out to make progress. Fix the lack
of wake-up.
We were also mis-handling the wait_event_timeout() return value - this
returns an unsigned integer of the remaining time, or zero on timeout
and the condition evaluated false. Checking this for less than zero
is not sane.
Signed-off-by: Russell King <[email protected]>
Even with the oops fixed by a previous patch, the system still fails to
kexec, due to a stuck chained interrupt locking the system. We must
disable the child interrupts prior to setting up the irq chip to ensure
we don't get stuck here.
Signed-off-by: Russell King <[email protected]>
Signed-off-by: Philipp Zabel <[email protected]>
We don't want GCC optimising our memset_io(), memcpy_fromio() or
memcpy_toio() variants, so we must not call one of the standard
functions. Provide a separate name for our assembly memcpy() and
memset() functions, and use that instead, thereby bypassing GCC's
ability to optimise these operations.
GCCs optimisation may introduce unaligned accesses which are invalid
for device mappings.
Signed-off-by: Russell King <[email protected]>
Convert the ioremap*() preprocessor macros to real functions, moving
them out of line. This allows us to kill off __arm_ioremap(), and
__arm_iounmap() helpers, and remove __arm_ioremap_pfn_caller() from
global view.
Signed-off-by: Russell King <[email protected]>
ioremap_wt() was added by aliasing it to ioremap_nocache(), which is a
device mapping. Device mappings do not allow unaligned accesses, but
it appears that GCC is able to inline its own memcpy() implementation
which may use such accesses. The only user of this is pmem, which
uses memcpy() on the region.
Therefore, this is unsafe. We must implement ioremap_wt() correctly
for ARM, or not at all.
This patch adds a more correct implementation by re-using ioremap_wc()
to provide a normal-memory non-cacheable mapping.
Signed-off-by: Russell King <[email protected]>
Add documentation of the ARM specific behaviour of the mappings setup by
the ioremap() series of macros.
Signed-off-by: Russell King <[email protected]>
Wolfram Sang reported an unannotated irqs-off warning from lockdep:
WARNING: CPU: 0 PID: 282 at kernel/locking/lockdep.c:3557 check_flags+0x84/0x1f4()
DEBUG_LOCKS_WARN_ON(current->hardirqs_enabled)
CPU: 0 PID: 282 Comm: rcS Tainted: G W 4.1.0-00002-g5b076054611833 #179
Hardware name: Generic Emma Mobile EV2 (Flattened Device Tree)
Backtrace:
[<c0012c94>] (dump_backtrace) from [<c0012e3c>] (show_stack+0x18/0x1c)
r6:c02dcc67 r5:00000009 r4:00000000 r3:00400000
[<c0012e24>] (show_stack) from [<c02510c8>] (dump_stack+0x20/0x28)
[<c02510a8>] (dump_stack) from [<c0022c44>] (warn_slowpath_common+0x8c/0xb4)
[<c0022bb8>] (warn_slowpath_common) from [<c0022cd8>] (warn_slowpath_fmt+0x38/0x40)
r8:c780f470 r7:00000000 r6:00000000 r5:c03b0570 r4:c0b7ec04
[<c0022ca4>] (warn_slowpath_fmt) from [<c004cd38>] (check_flags+0x84/0x1f4)
r3:c02e13d8 r2:c02dceaa
[<c004ccb4>] (check_flags) from [<c0050e50>] (lock_acquire+0x4c/0xbc)
r5:00000000 r4:60000193
[<c0050e04>] (lock_acquire) from [<c0256000>] (_raw_spin_lock+0x34/0x44)
r9:000a8d5c r8:00000001 r7:c7806000 r6:c780f460 r5:c03b06a0 r4:c780f460
[<c0255fcc>] (_raw_spin_lock) from [<c005a8cc>] (handle_fasteoi_irq+0x20/0x11c)
r4:c780f400
[<c005a8ac>] (handle_fasteoi_irq) from [<c0057a4c>] (generic_handle_irq+0x28/0x38)
r6:00000000 r5:c03b038c r4:00000012 r3:c005a8ac
[<c0057a24>] (generic_handle_irq) from [<c0057ae4>] (__handle_domain_irq+0x88/0xa8)
r4:00000000 r3:00000026
[<c0057a5c>] (__handle_domain_irq) from [<c000a3cc>] (gic_handle_irq+0x40/0x58)
r8:10c5347d r7:10c5347d r6:c35b1fb0 r5:c03a6304 r4:c8802000 r3:c35b1fb0
[<c000a38c>] (gic_handle_irq) from [<c0013bc8>] (__irq_usr+0x48/0x60)
Exception stack(0xc35b1fb0 to 0xc35b1ff8)
1fa0: 00000061 00000000 000ab736 00000066
1fc0: 00000061 000aa1f0 000a8d54 000a8d54 000a8d88 000a8d5c 000a8cc8 000a8d68
1fe0: 72727272 bef8a528 000398c0 00031334 20000010 ffffffff
r6:ffffffff r5:20000010 r4:00031334 r3:00000061
---[ end trace cb88537fdc8fa202 ]---
possible reason: unannotated irqs-off.
irq event stamp: 769
hardirqs last enabled at (769): [<c000f82c>] ret_fast_syscall+0x2c/0x54
hardirqs last disabled at (768): [<c000f80c>] ret_fast_syscall+0xc/0x54
softirqs last enabled at (0): [<c0020ec4>] copy_process.part.65+0x2e8/0x11dc
softirqs last disabled at (0): [< (null)>] (null)
His kernel configuration had:
CONFIG_PROVE_LOCKING=y
CONFIG_TRACE_IRQFLAGS=y
but no IRQSOFF_TRACER, which means entry from userspace can result in the
kernel seeing IRQs off without being notified of that change of state.
Change the IRQSOFF ifdef in the usr_entry macro to TRACE_IRQFLAGS instead.
Tested-by: Wolfram Sang <[email protected]>
Signed-off-by: Russell King <[email protected]>
Add a nmessage to suggest that HIGHMEM is enabled when physical memory
is truncated due to lack of virtual address space to map it in the low
memory mapping.
Signed-off-by: Russell King <[email protected]>
Add some help text for the HIGHPTE configuration entry. This comes from
the x86 entry, but reworded to be more a more accurate description of
what this option does.
Signed-off-by: Russell King <[email protected]>
randconfig testing reveals that DEBUG_SET_MODULE_RONX needs to depend on
MMU otherwise these build errors are observed:
kernel/built-in.o: In function `set_section_ro_nx':
kernel/module.c:1738: undefined reference to `set_memory_nx'
kernel/built-in.o: In function `set_page_attributes':
kernel/module.c:1709: undefined reference to `set_memory_ro'
This is because the pageattr functions are not built for !MMU configs as
they don't have page tables.
Signed-off-by: Russell King <[email protected]>
v3.18 changed handle_IRQ() to call __handle_domain_irq(), which now
rejects attempts to deliver IRQ0. Since IRQ 0 is used as the timer
interrupt (just like the PIT on x86), this causes boot to fail as the
bogomips calibration never completes.
Fix this by shuffling all interrupts up by one.
Fixes: a71b092a9c ("ARM: Convert handle_IRQ to use __handle_domain_irq")
Signed-off-by: Russell King <[email protected]>
Signed-off-by: Gregory CLEMENT <[email protected]>
The IPU code was installing its chained interrupt handler (which enables
the interrupt) before it was setting its data, which provokes an oops on
kexec. Fix this by converting to irq_set_chained_handler_and_data().
[drm] Initialized drm 1.1.0 20060810
imx-drm display-subsystem: parent device of /soc/aips-bus@02000000/ldb@020e0008/lvds-channel@1 is not available
imx-drm display-subsystem: parent device of /soc/aips-bus@02000000/ldb@020e0008/lvds-channel@1 is not available
Unable to handle kernel NULL pointer dereference at virtual address 00000070
pgd = c0004000
[00000070] *pgd=00000000
Internal error: Oops: 5 [#1] SMP ARM
Modules linked in:
CPU: 0 PID: 1 Comm: swapper/0 Not tainted 4.1.0-rc6+ #1693
Hardware name: Freescale i.MX6 Quad/DualLite (Device Tree)
task: d74c0000 ti: d74aa000 task.ti: d74aa000
PC is at ipu_irq_handle+0x28/0xd8
LR is at ipu_irq_handler+0x6c/0xc0
pc : [<c03c56d8>] lr : [<c03c58a4>] psr: 200001d3
sp : d74abbd0 ip : d74abc00 fp : d74abbfc
r10: 000001e0 r9 : c0085154 r8 : 00000009
r7 : 00000000 r6 : 00000000 r5 : d74abc04 r4 : c0a6b6a8
r3 : 00000000 r2 : 00000009 r1 : d74abc04 r0 : 00000000
Flags: nzCv IRQs off FIQs off Mode SVC_32 ISA ARM Segment kernel
Control: 10c5387d Table: 10004059 DAC: 00000015
Process swapper/0 (pid: 1, stack limit = 0xd74aa210)
Stack: (0xd74abbd0 to 0xd74ac000)
Backtrace:
[<c03c56b0>] (ipu_irq_handle) from [<c03c58a4>] (ipu_irq_handler+0x6c/0xc0)
[<c03c5838>] (ipu_irq_handler) from [<c0080154>] (generic_handle_irq+0x28/0x38)
[<c008012c>] (generic_handle_irq) from [<c0080288>] (__handle_domain_irq+0x5c/0xb8)
[<c008022c>] (__handle_domain_irq) from [<c0009428>] (gic_handle_irq+0x28/0x68)
[<c0009400>] (gic_handle_irq) from [<c0013dc4>] (__irq_svc+0x44/0x5c)
[<c07638fc>] (_raw_spin_unlock_irqrestore) from [<c00803bc>] (__irq_put_desc_unlock+0x1c/0x40)
[<c00803a0>] (__irq_put_desc_unlock) from [<c00841f4>] (__irq_set_handler+0x54/0x5c)
[<c00841a0>] (__irq_set_handler) from [<c03c5f48>] (ipu_probe+0x29c/0x708)
[<c03c5cac>] (ipu_probe) from [<c03d3848>] (platform_drv_probe+0x50/0xac)
[<c03d37f8>] (platform_drv_probe) from [<c03d1f3c>] (driver_probe_device+0x1d4/0x278)
Signed-off-by: Russell King <[email protected]>
Cc: Alexandre Courbot <[email protected]>
Cc: Hans Ulli Kroll <[email protected]>
Cc: Jason Cooper <[email protected]>
Cc: Lee Jones <[email protected]>
Cc: Linus Walleij <[email protected]>
Cc: Thierry Reding <[email protected]>
Cc: [email protected]
Link: http://lkml.kernel.org/r/[email protected]
Signed-off-by: Thomas Gleixner <[email protected]>
Driver authors seem to get the ordering of irq_set_chained_handler()
and irq_set_handler_data() wrong - ordering the former before the
latter. This opens a race window where, if there is an interrupt
pending, the handler will be called between these two calls,
potentially resulting in an oops.
Provide a single interface to set both of these together, especially
as that's commonly what is required.
Signed-off-by: Russell King <[email protected]>
Cc: Alexandre Courbot <[email protected]>
Cc: Hans Ulli Kroll <[email protected]>
Cc: Jason Cooper <[email protected]>
Cc: Lee Jones <[email protected]>
Cc: Linus Walleij <[email protected]>
Cc: Thierry Reding <[email protected]>
Cc: [email protected]
Link: http://lkml.kernel.org/r/[email protected]
Signed-off-by: Thomas Gleixner <[email protected]>
Commit 32e55a777f ("ARM: 8389/1: Add cpu_resume_arm() for firmwares
that resume in ARM state") needed to introduce a new usage of BSYM()
to fix a problem with a previous patch. This in turn causes a conflict
with the "bsym" branch which removes this symbol, replacing it with a
'badr' assembly macro. Fix this up.
Signed-off-by: Russell King <[email protected]>
Fix:
arch/arm/kernel/sleep.S:121: Error: selected processor does not support ARM opcodes
arch/arm/kernel/sleep.S:123: Error: attempt to use an ARM instruction on a Thumb-only processor -- `adr r9,1f+1'
arch/arm/kernel/sleep.S:124: Error: attempt to use an ARM instruction on a Thumb-only processor -- `bx r9'
Signed-off-by: Russell King <[email protected]>
There's quite a lot here, most of it from Mark Rutland, who has been
working on big.LITTLE PMU support for a while now. His work also brings
us significantly closer to moving the bulk of the CPU PMU driver out
into drivers/, where it can be shared with arm64.
As part of this work, there is a small patch to perf/core, which has
been Acked-by PeterZ and doesn't conflict with tip/perf/core at present.
I've kept that patch on a separate branch, merged in here, so that the
tip guys can pull it too if any unexpected issues crop up.
Please note that there is a conflict with mainline, since we remove
perf_event_cpu.c. The correct resolution is also to remove the file,
since the changes there are already reflected in the rework (and this
resolution is already included in linux-next).
Writing to a file is supposed to return the number of bytes written.
Returning zero unfortunately causes bash to constantly spin trying
to write to the sysfs file, to such an extent that even ^c and ^z
have no effect. The only way out of that is to kill the shell and
log back in. This isn't nice behaviour.
Fix it by returning the number of characters written to sysfs files.
[airlied: used suggestion from Al Viro]
Signed-off-by: Russell King <[email protected]>
Reviewed-by: Chris Wilson <[email protected]>
Signed-off-by: Dave Airlie <[email protected]>
A recent change in kernel/acct.c added a new warning for many
configurations on ARM:
kernel/acct.c: In function 'acct_pin_kill':
arch/arm/include/asm/cmpxchg.h:122:3: warning: value computed is not used [-Wunused-value]
The code is in fact correct, it's just a cmpxchg() call that
intentionally ignores the result, and no other code does that. The
warning does not show up on x86 because of the way that its cmpxchg()
macro is written. This changes the ARM implementation to use a similar
construct with a compound expression instead of a typecast, which causes
the compiler to not complain about an unused result.
Fix the other macros in this file in a similar way, and place them
just below their function implementations.
Signed-off-by: Russell King <[email protected]>
Document that r13 is not a stack in the initialisation function, in
case anyone gets other ideas.
Document the registers available for the errata workarounds, and
specifically which registers contain parts of the MIDR register, as
well as which registers must be preserved.
Lastly, use the lowest numbered available register (r0) rather than
r10 for temporary storage.
Signed-off-by: Russell King <[email protected]>
We already have the main ID register available in r9, there's no need
to refetch it. Use the saved value.
Signed-off-by: Russell King <[email protected]>
Rather than having a long sprawling __v7_setup function, which is hard
to maintain properly, move the CPU errata out of line.
While doing this, it was discovered that the Cortex-A15 errata had been
incorrectly added:
ldr r10, =0x00000c08 @ Cortex-A8 primary part number
teq r0, r10
bne 2f
/* Cortex-A8 errata */
b 3f
2: ldr r10, =0x00000c09 @ Cortex-A9 primary part number
teq r0, r10
bne 3f
/* Cortex-A9 errata */
3: ldr r10, =0x00000c0f @ Cortex-A15 primary part number
teq r0, r10
bne 4f
/* Cortex-A15 errata */
4:
This results in the Cortex-A15 test always being executed after the
Cortex-A8 and Cortex-A9 errata, which is obviously not what is intended.
The 'b 3f' labels should have been updated to 'b 4f'. The new structure
of:
/* Cortex-A8 Errata */
ldr r10, =0x00000c08 @ Cortex-A8 primary part number
teq r0, r10
beq __ca8_errata
/* Cortex-A9 Errata */
ldr r10, =0x00000c09 @ Cortex-A9 primary part number
teq r0, r10
beq __ca9_errata
/* Cortex-A15 Errata */
ldr r10, =0x00000c0f @ Cortex-A15 primary part number
teq r0, r10
beq __ca15_errata
__errata_finish:
is much cleaner and easier to see that this kind of thing doesn't
happen.
Signed-off-by: Russell King <[email protected]>
Re-engineer the LPAE TTBR setup code. Rather than passing some shifted
address in order to fit in a CPU register, pass either a full physical
address (in the case of r4, r5 for TTBR0) or a PFN (for TTBR1).
This removes the ARCH_PGD_SHIFT hack, and the last dangerous user of
cpu_set_ttbr() in the secondary CPU startup code path (which was there
to re-set TTBR1 to the appropriate high physical address space on
Keystone2.)
Tested-by: Murali Karicheri <[email protected]>
Signed-off-by: Russell King <[email protected]>
Eliminate the needless nommu version of this function, and get rid of
the proc_info_list structure argument - we no longer need this in order
to fix up the page table entries.
Acked-by: Santosh Shilimkar <[email protected]>
Tested-by: Murali Karicheri <[email protected]>
Signed-off-by: Russell King <[email protected]>
Re-implement the physical address space switching to be architecturally
compliant. This involves flushing the caches, disabling the MMU, and
only then updating the page tables. Once that is complete, the system
can be brought back up again.
Since we disable the MMU, we need to do the update in assembly code.
Luckily, the entries which need updating are fairly trivial, and are
all setup by the early assembly code. We can merely adjust each entry
by the delta required.
Not only does this fix the code to be architecturally compliant, but it
fixes a couple of bugs too:
1. The original code would only ever update the first L2 entry covering
a fraction of the kernel; the remainder were left untouched.
2. The L2 entries covering the DTB blob were likewise untouched.
This solution fixes up all entries.
Tested-by: Murali Karicheri <[email protected]>
Signed-off-by: Russell King <[email protected]>
The init_meminfo() method is not about initialising meminfo - it's about
fixing up the physical to virtual translation so that we use a different
physical address space, possibly above the 4GB physical address space.
Therefore, the name "init_meminfo()" is confusing.
Rename it to pv_fixup() instead.
Acked-by: Santosh Shilimkar <[email protected]>
Tested-by: Murali Karicheri <[email protected]>
Signed-off-by: Russell King <[email protected]>
There is no point platform code doing this, let's move it into the
generic code so it doesn't get duplicated.
Acked-by: Santosh Shilimkar <[email protected]>
Tested-by: Murali Karicheri <[email protected]>
Signed-off-by: Russell King <[email protected]>
Make the init_meminfo function return the offset to be applied to the
phys-to-virt translation constants. This allows us to move the update
into generic code, along with the requirements for this update.
This avoids platforms having to know the details of the phys-to-virt
translation support.
Acked-by: Santosh Shilimkar <[email protected]>
Tested-by: Murali Karicheri <[email protected]>
Signed-off-by: Russell King <[email protected]>
We ideally want the init_meminfo function to do nothing but return the
delta to be applied to PHYS_OFFSET - it should do nothing else. As we
can detect in platform init code whether we are running in the high
physical address space, move the platform notifier initialisation
entirely into the platform init code.
Acked-by: Santosh Shilimkar <[email protected]>
Tested-by: Murali Karicheri <[email protected]>
Signed-off-by: Russell King <[email protected]>
All ARMv5 and older CPUs invalidate their caches in the early assembly
setup function, prior to enabling the MMU. This is because the L1
cache should not contain any data relevant to the execution of the
kernel at this point; all data should have been flushed out to memory.
This requirement should also be true for ARMv6 and ARMv7 CPUs - indeed,
these typically do not search their caches when caching is disabled (as
it needs to be when the MMU is disabled) so this change should be safe.
ARMv7 allows there to be CPUs which search their caches while caching is
disabled, and it's permitted that the cache is uninitialised at boot;
for these, the architecture reference manual requires that an
implementation specific code sequence is used immediately after reset
to ensure that the cache is placed into a sane state. Such
functionality is definitely outside the remit of the Linux kernel, and
must be done by the SoC's firmware before _any_ CPU gets to the Linux
kernel.
Changing the data cache clean+invalidate to a mere invalidate allows us
to get rid of a lot of platform specific hacks around this issue for
their secondary CPU bringup paths - some of which were buggy.
Reviewed-by: Florian Fainelli <[email protected]>
Tested-by: Florian Fainelli <[email protected]>
Tested-by: Heiko Stuebner <[email protected]>
Tested-by: Dinh Nguyen <[email protected]>
Acked-by: Sebastian Hesselbarth <[email protected]>
Tested-by: Sebastian Hesselbarth <[email protected]>
Acked-by: Shawn Guo <[email protected]>
Tested-by: Thierry Reding <[email protected]>
Acked-by: Thierry Reding <[email protected]>
Tested-by: Geert Uytterhoeven <[email protected]>
Tested-by: Michal Simek <[email protected]>
Tested-by: Wei Xu <[email protected]>
Signed-off-by: Russell King <[email protected]>
The drm_mm debugfs output is difficult to read as two different formats
are used for the addresses:
0x00000080000000-0x0000008000b000: 45056: used
0x8000b000-0x80016000: 45056: free
0x00000080016000-0x0000008001b000: 20480: used
0x8001b000-0x817a1000: 24666112: free
0x000000817a1000-0x000000817a8000: 28672: used
0x000000817a8000-0x00000081ba8000: 4194304: used
Fix this by using %#018llx for all addresses, thus making the output:
0x0000000080000000-0x000000008000b000: 45056: used
0x000000008000b000-0x0000000080016000: 45056: free
0x0000000080016000-0x000000008001b000: 20480: used
0x000000008001b000-0x00000000817a1000: 24666112: free
0x00000000817a1000-0x00000000817a8000: 28672: used
0x00000000817a8000-0x0000000081ba8000: 4194304: used
Signed-off-by: Russell King <[email protected]>
Reviewed-by: Alex Deucher <[email protected]>
Signed-off-by: Dave Airlie <[email protected]>
Add a helper to create the IEC958 channel status from an ALSA
snd_pcm_runtime structure, taking account of the sample rate and
sample size.
Signed-off-by: Russell King <[email protected]>
Reviwed-by: Mark Brown <[email protected]>
Signed-off-by: Takashi Iwai <[email protected]>
Add a helper for the EDID like data structure, which is typically passed
from a HDMI adapter to its associated audio driver. This informs the
audio driver of the capabilities of the attached HDMI sink.
Signed-off-by: Russell King <[email protected]>
Reviewed-by: Mark Brown <[email protected]>
Signed-off-by: Takashi Iwai <[email protected]>
Stephen Rothwell reports that he sees a compiler warning on x86_64:
drivers/gpu/drm/i2c/tda998x_drv.c: In function 'tda998x_write_avi':
drivers/gpu/drm/i2c/tda998x_drv.c:647:3: warning: format '%d' expects argument of type 'int', but argument 3 has type 'ssize_t' [-Wformat=]
dev_err(&priv->hdmi->dev, "hdmi_avi_infoframe_pack() failed: %d\n", len);
^
Fix this by using the appropriate length modifier.
Signed-off-by: Russell King <[email protected]>
Avoid passing the auxiliary control register value through the enable
method. In the resume path, we have to read the value stored in
l2x0_saved_regs.aux_ctrl, only to have it immediately written back by
l2c_enable(). We can avoid this if we have __l2c_init() save the value
directly to l2x0_saved_regs.aux_ctrl before calling the specific enable
method.
Signed-off-by: Russell King <[email protected]>
Some L2C caches have a bit which allows non-secure software to control
the cache lockdown. Some platforms are unable to set this bit. To
avoid receiving an abort while trying to unlock the cache lines, check
the state of this bit before unlocking. We do this by providing a new
method in the l2c_init_data to perform the unlocking.
Signed-off-by: Russell King <[email protected]>
l2c_configure() does not follow the pattern of other l2c_* functions.
Fix this so that it does to avoid future confusion.
Signed-off-by: Russell King <[email protected]>
Before calling the controller specific configuration function, write
the auxiliary control register first, so that bits shared with other
registers (such as the prefetch control register) are not overwritten
by the later write to the auxctrl register.
Signed-off-by: Russell King <[email protected]>
l2c_enable() is documented that it must not be called if the cache has
already been enabled. Unfortunately, commit 6b49241ac2 ("ARM: 8259/1:
l2c: Refactor the driver to use commit-like interface") changed this
without updating the comment, for very little reason. Revert this
change and restore the expected behaviour.
Signed-off-by: Russell King <[email protected]>
Josh Stone reports:
I've discovered a case where both arm and arm64 will miss a ptrace
syscall-exit that they should report. If the syscall is entered
without TIF_SYSCALL_TRACE set, then it goes on the fast path. It's
then possible to have TIF_SYSCALL_TRACE added in the middle of the
syscall, but ret_fast_syscall doesn't check this flag again.
Fix this by always checking for a syscall trace in the fast exit path.
Reported-by: Josh Stone <[email protected]>
Signed-off-by: Russell King <[email protected]>
The SI unit of frequency is Hertz, named after Heinrich Hertz, and is
given the symbol "Hz" to denote this. "hz" is not the unit of frequency,
and is in fact meaningless.
Fix arch/arm to correctly use "Hz", thereby acknowledging Heinrich Hertz'
contribution to the modern world.
Acked-by: Tony Lindgren <[email protected]>
Acked-by: Robert Jarzmik <[email protected]>
Reviewed-by: Andreas Färber <[email protected]>
Signed-off-by: Russell King <[email protected]>
BSYM() was invented to allow us to work around a problem with the
assembler, where local symbols resolved by the assembler for the 'adr'
instruction did not take account of their ISA.
Since we don't want BSYM() used elsewhere, replace BSYM() with a new
macro 'badr', which is like the 'adr' pseudo-op, but with the BSYM()
mechanics integrated into it. This ensures that the BSYM()-ification
is only used in conjunction with 'adr'.
Acked-by: Dave Martin <[email protected]>
Acked-by: Nicolas Pitre <[email protected]>
Signed-off-by: Russell King <[email protected]>
BSYM() should only be used when refering to local symbols in the same
assembly file which are resolved by the assembler, and not for
linker-fixed up symbols. The use of BSYM() with panic is incorrect as
the linker is involved in fixing up this relocation, and it knows
whether panic() is ARM or Thumb.
Acked-by: Nicolas Pitre <[email protected]>
Acked-by: Dave Martin <[email protected]>
Signed-off-by: Russell King <[email protected]>
If we are building for a LE platform, and we haven't overriden the
MMIO ops, then we can optimize the mem*io operations using the
standard string functions.
Acked-by: Nicolas Pitre <[email protected]>
Signed-off-by: Russell King <[email protected]>
clkdev_create() is a shorter way to write clkdev_alloc() followed by
clkdev_add(). Use this instead.
Acked-by: Mark Brown <[email protected]>
Signed-off-by: Russell King <[email protected]>
When creating aliases of existing clkdev clocks, use clkdev_add_alias()
isntead of open coding the lookup and clk_lookup creation.
Acked-by: Tony Lindgren <[email protected]>
Signed-off-by: Russell King <[email protected]>
Rather than open coding the clkdev allocation, initialisation and
addition, use the clkdev_create() helper.
Acked-by: Tony Lindgren <[email protected]>
Signed-off-by: Russell King <[email protected]>
clkdev_create() is a shorter way to write clkdev_alloc() followed by
clkdev_add(). Use this instead.
Acked-by: Andrew Lunn <[email protected]>
Signed-off-by: Russell King <[email protected]>
We have always had an efficient way of registering a table of clock
lookups - it's called clkdev_add_table(). However, some people seem
to really love writing inefficient and unnecessary code.
Convert LPC32xx to use the correct interface.
Signed-off-by: Russell King <[email protected]>
We have always had an efficient way of registering a table of clock
lookups - it's called clkdev_add_table(). However, some people seem
to really love writing inefficient and unnecessary code.
Convert SH to use the correct interface.
Acked-by: Geert Uytterhoeven <[email protected]>
Signed-off-by: Russell King <[email protected]>
Add a helper to allocate and add a clk_lookup structure. This can not
only be used in several places in clkdev.c to simplify the code, but
more importantly, can be used by callers of the clkdev code to simplify
their clkdev creation and registration.
Signed-off-by: Russell King <[email protected]>
clk_add_alias() is provided by clkdev, and is not part of the clk API.
Howver, it is prototyped in two locations: linux/clkdev.h and
linux/clk.h. This is a mess. Get rid of the redundant and unnecessary
version in linux/clk.h.
Acked-by: Tony Lindgren <[email protected]>
Tested-by: Robert Jarzmik <[email protected]>
Acked-by: Sekhar Nori <[email protected]>
Signed-off-by: Russell King <[email protected]>
We want to be able to call clkdev_add_table() from non-init code, so we
need to drop the __init marker from it.
Signed-off-by: Russell King <[email protected]>
The idea is that rate = clk_round_rate(clk, r) is equivalent to:
clk_set_rate(clk, r);
rate = clk_get_rate(clk);
except that clk_round_rate() does not change the hardware in any way.
Signed-off-by: Russell King <[email protected]>
clk_add_alias() calls clk_get() followed by clk_put() but in between
those two calls it saves away the struct clk pointer to a clk_lookup
structure. This leaves the 'clk' member of the clk_lookup pointing at
freed memory on configurations where CONFIG_COMMON_CLK=y. This is a
problem because clk_get_sys() will eventually try to dereference the
freed pointer by calling __clk_get_hw() on it. Fix this by saving away
the struct clk_hw pointer instead of the struct clk pointer so that when
we try to create a per-user struct clk in clk_get_sys() we don't
dereference a junk pointer.
Signed-off-by: Russell King <[email protected]>
Commit bf35706f3d ("ARM: 8314/1: replace PROCINFO embedded branch with
relative offset") broke booting on nommu platforms as it didn't update
the nommu boot code. This patch fixes that oversight.
Fixes: bf35706f3d ("ARM: 8314/1: replace PROCINFO embedded branch with relative offset")
Signed-off-by: Russell King <[email protected]>
This errata covers all r1 variants of Cortex A8, it's not limited to
just r1p0..r1p2. Update the documentation to reflect this. The code
already applies the workaround to all r1p* A8 CPUs.
Signed-off-by: Russell King <[email protected]>
We have recently had an example of someone wanting to use a 90kHz timer
for the software delay loop.
udelay() needs to have at least microsecond resolution to allow drivers
access to a delay mechanism with a reasonable chance of delaying the
period they requested within at least a 50% marging of error, especially
for small delays.
Discussion about the udelay() accuracy can be found at:
https://lkml.org/lkml/2011/1/9/37
Reject timers which are unable to supply this level of resolution.
Acked-by: Nicolas Pitre <[email protected]>
Signed-off-by: Russell King <[email protected]>
Switch ARM to use the generic show_mem() implementation, which displays
the statistics from the mm zone rather than walking the page arrays.
Acked-by: Mel Gorman <mgorman <[email protected]>
Tested-by: Gregory Fong <[email protected]>
Signed-off-by: Russell King <[email protected]>
Avoid the errata 430973 workaround for non-Cortex A8 CPUs. Having this
workaround enabled introduces an additional branch target buffer flush
into the context switching path, something we wish to avoid. To allow
this errata to be enabled in multiplatform kernels while reducing its
impact, rearrange the Cortex-A8 CPU support to avoid impacting on other
Version 7 CPUs.
Tested-by: Tony Lindgren <[email protected]>
Signed-off-by: Russell King <[email protected]>
The effects of not having ARM errata 643719 enabled on affected CPUs
can be very confusing and hard to debug. Rather than leave this to
chance, enable this workaround by default. Now that we have rearranged
the code, it should have a low impact on the majority of CPUs.
Acked-by: Catalin Marinas <[email protected]>
Reviewed-by: Catalin Marinas <[email protected]>
Signed-off-by: Russell King <[email protected]>
Eliminate one unnecessary instruction from this test by pre-shifting
the Cortex A9 ID - we can shift the actual ID in the teq instruction
thereby losing the pX bit of the ID at no cost.
Reviewed-by: Catalin Marinas <[email protected]>
Signed-off-by: Russell King <[email protected]>
Optimise the branches such that for the majority of unaffected devices,
we avoid needing to execute the errata work-around code path by
branching to start_flush_levels early.
Reviewed-by: Catalin Marinas <[email protected]>
Signed-off-by: Russell King <[email protected]>
Both v7_flush_cache_louis and v7_flush_dcache_all both begin the
flush_levels loop with r10 initialised to zero. In each case, this
is done immediately prior to entering the loop. Branch to this
instruction in v7_flush_dcache_all from v7_flush_cache_louis and
eliminate the unnecessary initialisation in v7_flush_cache_louis.
Reviewed-by: Catalin Marinas <[email protected]>
Signed-off-by: Russell King <[email protected]>
Rather than have code which masks and then shifts, such as:
mrc p15, 1, r0, c0, c0, 1
ALT_SMP(ands r3, r0, #7 << 21)
ALT_UP( ands r3, r0, #7 << 27)
ALT_SMP(mov r3, r3, lsr #20)
ALT_UP( mov r3, r3, lsr #26)
re-arrange this as a shift and then mask. The masking is the same for
each field which we want to extract, so this allows the mask to be
shared amongst code paths:
mrc p15, 1, r0, c0, c0, 1
ALT_SMP(mov r3, r0, lsr #20)
ALT_UP( mov r3, r0, lsr #26)
ands r3, r3, #7 << 1
Use this method for the LoUIS, LoUU and LoC fields.
Reviewed-by: Catalin Marinas <[email protected]>
Signed-off-by: Russell King <[email protected]>
We always build cache-v7.S for ARMv7, so we can use the ARMv7 16-bit
move instructions to load large constants, rather than using constants
in a literal pool.
Reviewed-by: Catalin Marinas <[email protected]>
Signed-off-by: Russell King <[email protected]>
Allow ALT_UP() to cope with a 16-bit Thumb instruction by automatically
inserting a following nop instruction. This allows us to care less
about getting the assembler to emit a 32-bit thumb instruction.
Signed-off-by: Russell King <[email protected]>
No merged platform supplies xclks via platform data. As we want to
slightly change the clkdev interface, rather than fixing this unused
code, remove it instead.
Signed-off-by: Russell King <[email protected]>
Signed-off-by: Laurent Pinchart <[email protected]>
Acked-by: Sakari Ailus <[email protected]>
Signed-off-by: Mauro Carvalho Chehab <[email protected]>
Both ARM946 and ARM940 setup functions were corrupting r1 and r2,
which is not permissible - these are used to carry the machine ID
and boot data into the kernel, and must be preserved.
The code responsible for this was the same in both files: they were
using the registers to generate a protection region register value.
Fix this by turning this process into a macro, and using that macro
in both these files with an alternative register allocation. r0,
r3 and r7 can be used for temporary values here.
Reported-by: Alex Dumitrache <[email protected]>
Tested-by: Georg Hofstetter <[email protected]>
Signed-off-by: Russell King <[email protected]>
The Cortex A9 tile fails to unplug CPUs if errata 643719 is not enabled.
This leads to random weird behaviours, but ultimately seem to lock the
kernel one way or another when a CPU is hot unplugged.
Symptoms range from a spinlock lockup in the scheduler, the entire
system hanging, to dumping out the kernel printk buffer a few lines at
a time, and other weird behaviours.
This is caused by the outgoing CPU not having its inner caches properly
flushed before it exits coherency - flush_cache_louis() is used to
achieve this, but as a result of the hardware bug, this function ends
up doing nothing without the errata workaround enabled.
As the Versatile Express has an affected CPU, this errata must always
be enabled.
Signed-off-by: Russell King <[email protected]>
Move shutdown and reboot related code to a separate file, out of
process.c. This helps to avoid polluting process.c with non-process
related code.
Signed-off-by: Russell King <[email protected]>
Normally, when a CPU wants to clear a cache line to zero in the external
L2 cache, it would generate bus cycles to write each word as it would do
with any other data access.
However, a Cortex A9 connected to a L2C-310 has a specific feature where
the CPU can detect this operation, and signal that it wants to zero an
entire cache line. This feature, known as Full Line of Zeros (FLZ),
involves a non-standard AXI signalling mechanism which only the L2C-310
can properly interpret.
There are separate enable bits in both the L2C-310 and the Cortex A9 -
the L2C-310 needs to be enabled and have the FLZ enable bit set in the
auxiliary control register before the Cortex A9 has this feature
enabled.
Unfortunately, the suspend code was not respecting this - it's not
obvious from the code:
swsusp_arch_suspend()
cpu_suspend() /* saves the Cortex A9 auxiliary control register */
arch_save_image()
soft_restart() /* turns off FLZ in Cortex A9, and disables L2C */
cpu_resume() /* restores the Cortex A9 registers, inc auxcr */
At this point, we end up with the L2C disabled, but the Cortex A9 with
FLZ enabled - which means any memset() or zeroing of a full cache line
will fail to take effect.
A similar issue exists in the resume path, but it's slightly more
complex:
swsusp_arch_suspend()
cpu_suspend() /* saves the Cortex A9 auxiliary control register */
arch_save_image() /* image with A9 auxcr saved */
...
swsusp_arch_resume()
call_with_stack()
arch_restore_image() /* restores image with A9 auxcr saved above */
soft_restart() /* turns off FLZ in Cortex A9, and disables L2C */
cpu_resume() /* restores the Cortex A9 registers, inc auxcr */
Again, here we end up with the L2C disabled, but Cortex A9 FLZ enabled.
There's no need to turn off the L2C in either of these two paths; there
are benefits from not doing so - for example, the page copies will be
faster with the L2C enabled.
Hence, fix this by providing a variant of soft_restart() which can be
used without turning the L2 cache controller off, and use it in both
of these paths to keep the L2C enabled across the respective resume
transitions.
Fixes: 8ef418c717 ("ARM: l2c: trial at enabling some Cortex-A9 optimisations")
Reported-by: Sean Cross <[email protected]>
Tested-by: Sean Cross <[email protected]>
Signed-off-by: Russell King <[email protected]>
Patch derived from one from Yakir Yang. Yakir Yang says:
For Designerware HDMI, the following write sequence is recommended:
1. aud_n3 (set bit ncts_atomic_write if desired)
2. aud_cts3 (set CTS_manual and CTS value if desired/enabled)
3. aud_cts2 (required in CTS_manual)
4. aud_cts1 (required in CTS_manual)
5. aud_n3 (bit ncts_atomic_write with same value as in step 1.)
6. aud_n2
7. aud_n1
However, avoid the ncts_atomic_write_bit and CTS_manual settings in this
patch, both of which are marked reserved in the iMX6 documentation. All
iMX6 code in the wild seems to want CTS_manual cleared.
Having requested clarification from FSL, it appears that neither of
these bits are implemented in their version of the IP.
Acked-by: Philipp Zabel <[email protected]>
Tested-by: Philipp Zabel <[email protected]>
Acked-by: Andy Yan <[email protected]>
Signed-off-by: Russell King <[email protected]>
The HDMI n/cts settings need to be updated whenever the audio sample
rate or the video pixel clock changes. This needs to be protected
against concurrency as there is no synchronisation between these two
operations. Introduce a mutex (called audio_mutex) to protect against
two threads trying to update the video clock rate and pixel clock
simultaneously.
Acked-by: Philipp Zabel <[email protected]>
Tested-by: Philipp Zabel <[email protected]>
Acked-by: Andy Yan <[email protected]>
Signed-off-by: Russell King <[email protected]>
Russell King