The TLB flushing functions on hppa, which causes PxTLB broadcasts on the system
bus, needs to be protected by irq-safe spinlocks to avoid irq handlers to deadlock
the kernel. The deadlocks only happened during I/O intensive loads and triggered
pretty seldom, which is why this bug went so long unnoticed.
Signed-off-by: Helge Deller <deller@gmx.de>
[edited to use spin_lock_irqsave on UP as well since we'd been locking there
all this time anyway, --kyle]
Signed-off-by: Kyle McMartin <kyle@mcmartin.ca>
Rewrote timer_interrupt() to properly handle the "delayed!" case.
If we used floating point math to compute the number of ticks that had
elapsed since the last timer interrupt, it could take up to 12K cycles
(emperical!) to handle the interrupt. Existing code assumed it would
never take more than 8k cycles. We end up programming Interval Timer
to a value less than "current" cycle counter. Thus have to wait until
Interval Timer "wrapped" and would then get the "delayed!" printk that
I moved below.
Since we don't really know what the upper limit is, I prefer to read
CR16 again after we've programmed it to make sure we won't have to
wait for CR16 to wrap.
Further, the printk was between reading CR16 (cycle couner) and writing CR16
(the interval timer). This would cause us to continue to set the interval
timer to a value that was "behind" the cycle counter. Rinse and repeat.
So no printk's between reading CR16 and setting next interval timer.
Tested on A500 (550 Mhz PA8600).
Signed-off-by: Grant Grundler <grundler@parisc-linux.org>
Tested-by: Kyle McMartin <kyle@mcmartin.ca>
Signed-off-by: Kyle McMartin <kyle@mcmartin.ca>
----
Kyle, Helge, and other parisc's,
Please test on 32-bit before committing.
I think I have it right but recognize I might not.
TODO: I wanted to use "do_div()" in order to get both remainder
and value back with one division op. That should help with the
latency alot but can be applied seperately from this patch.
thanks,
grant
>>>> I think this is what was intended? Note that this patch may affect
>>>> profiling.
>>> it really should be
>>>
>>> - if (likely(t1 & (sizeof(unsigned int)-1)) == 0) {
>>> + if (likely((t1 & (sizeof(unsigned int)-1)) == 0)) {
>>>
>>> randolph
Reported-by: Roel Kluin <roel.kluin@gmail.com>
Signed-off-by: Randolph Chung <tausq@parisc-linux.org>
Signed-off-by: Kyle McMartin <kyle@mcmartin.ca>
gcc 4.4 warns about:
drivers/parisc/lba_pci.c: In function 'lba_pat_resources':
drivers/parisc/lba_pci.c:1099: warning: the frame size of 8280 bytes is larger than 4096 bytes
The problem is we declare two large structures on the stack. They don't need
to be on the stack since they are only used during LBA initialization (which
is serialized). Moving to be "static".
Signed-off-by: Grant Grundler <grundler@parisc-linux.org>
Signed-off-by: Kyle McMartin <kyle@mcmartin.ca>
The defines and typedefs (hw_interrupt_type, no_irq_type, irq_desc_t) have
been kept around for migration reasons. After more than two years it's
time to remove them finally.
This patch cleans up one of the remaining users. When all such patches
hit mainline we can remove the defines and typedefs finally.
Impact: cleanup
Convert the last remaining users to struct irq_chip and remove the
define.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Kyle McMartin <kyle@mcmartin.ca>
Fix miscompilation in arch/parisc/kernel/irq.c:
123: warning: passing arg 1 of `cpumask_setall' from incompatible pointer type
141: warning: passing arg 1 of `cpumask_copy' from incompatible pointer type
300: warning: passing arg 1 of `cpumask_copy' from incompatible pointer type
357: warning: passing arg 2 of `cpumask_copy' from incompatible pointer type
Signed-off-by: Helge Deller <deller@gmx.de>
Signed-off-by: Kyle McMartin <kyle@mcmartin.ca>
Alex Chiang asked me why PARISC was calling pci_bus_add_devices()
and pci_bus_assign_resources() in the opposite order from everyone else.
No reason and I couldn't see any data dependency.
Patch below applies cleanly to 2.6.30-rc2.
Later, I suspected the code worked only because no drivers would be
loaded/ready until much later in the system initialization sequence.
Tested "LBA" code on J6000 (32-bit) and A500 (64-bit SMP) with 2.6.30-rc2.
Not tested with any Dino controllers.
Not tested with PCI-PCI Bridge (TBD).
Reported-by: Alex Chiang <achiang@hp.com>
Signed-off-by: Grant Grundler <grundler@parisc-linux.org>
Signed-off-by: Kyle McMartin <kyle@mcmartin.ca>
There are two reasons to expose the memory *a in the asm:
1) To prevent the compiler from discarding a preceeding write to *a, and
2) to prevent it from caching *a in a register over the asm.
The change has had a few days testing with a SMP build of 2.6.22.19
running on a rp3440.
This patch is about the correctness of the __ldcw() macro itself.
The use of the macro should be confined to small inline functions
to try to limit the effect of clobbering memory on GCC's optimization
of loads and stores.
Signed-off-by: Dave Anglin <dave.anglin@nrc-cnrc.gc.ca>
Signed-off-by: Helge Deller <deller@gmx.de>
Signed-off-by: Kyle McMartin <kyle@mcmartin.ca>
Doing an IPI with local interrupts off triggers a warning. We
don't need to be quite so ridiculously paranoid. Also, clean up
a bit of the code a little.
Signed-off-by: Kyle McMartin <kyle@mcmartin.ca>
The atomic operations on parisc are defined as macros. The macros
includes casts which disallows the use of some syntax elements and
produces error like this:
net/phonet/pep.c: In function 'pipe_rcv_status':
net/phonet/pep.c:262: error: lvalue required as left operand of assignment
The patch removes this superfluous casts.
Signed-off-by: Bastian Blank <waldi@debian.org>
Signed-off-by: Kyle McMartin <kyle@mcmartin.ca>
Fix this build error when CONFIG_PROC_FS is not set:
drivers/parisc/ccio-dma.c:1574: error: 'ccio_proc_info_fops' undeclared
Signed-off-by: Alexander Beregalov <a.beregalov@gmail.com>
Signed-off-by: Kyle McMartin <kyle@mcmartin.ca>
Fix this build error when CONFIG_STI_CONSOLE is not set
drivers/video/stifb.c:1337: undefined reference to `sti_get_rom'
Signed-off-by: Alexander Beregalov <a.beregalov@gmail.com>
Signed-off-by: Kyle McMartin <kyle@mcmartin.ca>
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/vapier/blackfin: (27 commits)
Blackfin: fix dma-mapping build errors
Blackfin: hook up new perf_counter_open syscall
Blackfin: drop BF535-specific text for exception 0x2A (unaligned instruction)
Blackfin: fix early crash when booting on wrong cpu
Blackfin: fix GPTMR0_CLOCKSOURCE dependency on BFIN_GPTIMERS
Blackfin: drop unused ISP1760 port1_disable from board resources
Blackfin: bf526-ezbrd: handle different SDRAM chips
Blackfin: fix typo in TRAS define in mem_init.h header
Blackfin: unify memory map headers
Blackfin: stick the CPU name into boot image name
Blackfin: update defconfigs
Blackfin: decouple unrelated cache settings to get exact behavior
Blackfin: update I-pipe patch level
Blackfin: remove obsolete mcount support from I-pipe code
Blackfin: allow CONFIG_TICKSOURCE_GPTMR0 with interrupt pipeline
Blackfin: convert interrupt pipeline to irqflags
Blackfin: allow people to select BF51x-0.1 silicon rev
Blackfin: bf526-ezbrd: set SPI flash resources to SST device
Blackfin: fix accidental reset in some boot modes
Blackfin: abstract irq14 lowering in do_irq
...
* git://git.infradead.org/~dwmw2/iommu-2.6.31:
intel-iommu: Fix one last ia64 build problem in Pass Through Support
VT-d: support the device IOTLB
VT-d: cleanup iommu_flush_iotlb_psi and flush_unmaps
VT-d: add device IOTLB invalidation support
VT-d: parse ATSR in DMA Remapping Reporting Structure
PCI: handle Virtual Function ATS enabling
PCI: support the ATS capability
intel-iommu: dmar_set_interrupt return error value
intel-iommu: Tidy up iommu->gcmd handling
intel-iommu: Fix tiny theoretical race in write-buffer flush.
intel-iommu: Clean up handling of "caching mode" vs. IOTLB flushing.
intel-iommu: Clean up handling of "caching mode" vs. context flushing.
VT-d: fix invalid domain id for KVM context flush
Fix !CONFIG_DMAR build failure introduced by Intel IOMMU Pass Through Support
Intel IOMMU Pass Through Support
Fix up trivial conflicts in drivers/pci/{intel-iommu.c,intr_remapping.c}
As noted in the previous patch, the NFSv4 client mount code currently
has several limitations. If the mount path contains symlinks, or
referrals, or even if it just contains a '..', then the client code in
nfs4_path_walk() will fail with an error.
This patch replaces the nfs4_path_walk()-based lookup with a helper
function that sets up a private namespace to represent the namespace on the
server, then uses the ordinary VFS and NFS path lookup code to walk down the
mount path in that namespace.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The purpose of this patch is to improve the remote mount path lookup
support for distributed filesystems such as the NFSv4 client.
When given a mount command of the form "mount server:/foo/bar /mnt", the
NFSv4 client is required to look up the filehandle for "server:/", and
then look up each component of the remote mount path "foo/bar" in order
to find the directory that is actually going to be mounted on /mnt.
Following that remote mount path may involve following symlinks,
crossing server-side mount points and even following referrals to
filesystem volumes on other servers.
Since the standard VFS path lookup code already supports walking paths
that contain all these features (using in-kernel automounts for
following referrals) we would like to be able to reuse that rather than
duplicate the full path traversal functionality in the NFSv4 client code.
This patch therefore defines a VFS helper function create_mnt_ns(), that
sets up a temporary filesystem namespace and attaches a root filesystem to
it. It exports the create_mnt_ns() and put_mnt_ns() function for use by
filesystem modules.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
In order to allow modules to use it without having to export vfsmount_lock.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
SLAB uses get/put_online_cpus() which use a mutex which is itself only
initialized when cpu_hotplug_init() is called. Currently we hang suring
boot in SLAB due to doing that too late.
Reported by James Bottomley and Sachin Sant (and possibly others).
Debugged by Benjamin Herrenschmidt.
This just removes the dynamic initialization of the data structures, and
replaces it with a static one, avoiding this dependency entirely, and
removing one unnecessary special initcall.
Tested-by: Sachin Sant <sachinp@in.ibm.com>
Tested-by: James Bottomley <James.Bottomley@HansenPartnership.com>
Tested-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The recent deprecation of dma_sync_{sg,single} ironically broke Blackfin
systems. This is because we don't define dma_sync_sg_for_cpu at all, so
until the DMA asm-generic conversion/cleanup is done after the next
release, simply stub out the dma_sync_sg_for_{cpu,device} functions.
Signed-off-by: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
We don't support the BF535 at all, and the exception 0x2A text specific to
it is pretty verbose and confusing (since the behavior is simply odd), so
punt it to keep the noise down.
Signed-off-by: Yi Li <yi.li@analog.com>
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Make sure we process the kernel command line before poking the hardware,
so that we can process early printk. This helps ensure that if you boot
a kernel configured for a different processor, something will be left in
the log buffer.
Signed-off-by: Robin Getz <robin.getz@analog.com>
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
The GPTMR0_CLOCKSOURCE Kconfig option requires the gptimers framework, so
make sure it is selected when this option is enabled.
Reported-by: Peter Meerwald <pmeerw@pmeerw.net>
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
The BF526-EZBRD changed SDRAM chips between board revisions, so create a
timing table that can accommodate both.
Signed-off-by: Graf Yang <graf.yang@analog.com>
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
We defined SDRAM_tRAS to TRAS_4, but then wrongly defined SDRAM_tRAS_num
to 3.
Signed-off-by: Graf Yang <graf.yang@analog.com>
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Many aspects of the Blackfin memory map is exactly the same across all
variants. Rather than copy and paste all of these duplicated values in
each header, unify all of these into the common Blackfin memory map header
file. In the process, push down BF561 SMP specific stuff to the BF561
specific header to keep the noise down.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Rather than use "Linux" in the boot image name (as this is redundant --
the image type is already set to "linux"), use the CPU name. This makes
it fairly obvious when a wrong image is accidentally booted. Otherwise
there is no kernel output and you waste time scratching your head
wondering wtf just happened.
Signed-off-by: Robin Getz <rgetz@blackfin.uclinux.org>
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
The current cache options don't really represent the hardware features.
They end up setting different aspects of the hardware so that the end
result is to turn on/off the cache. Unfortunately, when we hit cache
problems with the hardware, it's difficult to test different settings to
root cause the problem. The current settings also don't cleanly allow for
different caching behaviors with different regions of memory.
So split the configure options such that they properly reflect the settings
that are applied to the hardware.
Signed-off-by: Jie Zhang <jie.zhang@analog.com>
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
The BF526-EZBRD has a SST SPI flash on it, not a ST Micro.
Signed-off-by: Graf Yang <graf.yang@analog.com>
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
We read the SWRST (Software Reset) register to get at the last reset
state, and then we may configure the DOUBLE_FAULT bit to control behavior
when a double fault occurs. But if the lower bits of the register is
already set (like UART boot mode on a BF54x), we inadvertently make the
system reset by writing to the SYSTEM_RESET field at the same time. So
make sure the lower 4 bits are always cleared.
Signed-off-by: Sonic Zhang <sonic.zhang@analog.com>
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Split out the optional IRQ14 lowering code to further simplify the
asm_do_IRQ() function and keep the ifdef nest under control.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Take a page from x86 and abstract the stack checking out of the
asm_do_IRQ() function so that the result is easier to digest.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
With the common IRQ code initializing much more of the irq_desc state, we
can't blindly initialize it ourselves to the local bad_irq state. If we
do, we end up wrongly clobbering many fields. So punt most of the bad irq
code as the common layers will handle the default state, and simply call
handle_bad_irq() directly when the IRQ we are processing is invalid.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
The BF533-EZKIT has two Flash In-System Programming devices hooked up to
the async memory bus, so add resources for the primary flashes and the
SRAMs on the devices.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Helge Deller