This is much the same as for SPARC except that we can do the find_zero()
function more efficiently using the count-leading-zeroes instructions.
Tested on 32-bit and 64-bit PowerPC.
Signed-off-by: Paul Mackerras <paulus@samba.org>
Acked-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Remove CONFIG_POWER4_ONLY, the option is badly named and only does two
things:
- It wraps the MMU segment table code. With feature fixups there is
little downside to compiling this in.
- It uses the newer mtocrf instruction in various assembly functions.
Instead of making this a compile option just do it at runtime via
a feature fixup.
Signed-off-by: Anton Blanchard <anton@samba.org>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
This is no longer selectable, so just remove all the dependent code.
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Implement a POWER7 optimised copy_to_user/copy_from_user using VMX.
For large aligned copies this new loop is over 10% faster, and for
large unaligned copies it is over 200% faster.
If we take a fault we fall back to the old version, this keeps
things relatively simple and easy to verify.
On POWER7 unaligned stores rarely slow down - they only flush when
a store crosses a 4KB page boundary. Furthermore this flush is
handled completely in hardware and should be 20-30 cycles.
Unaligned loads on the other hand flush much more often - whenever
crossing a 128 byte cache line, or a 32 byte sector if either sector
is an L1 miss.
Considering this information we really want to get the loads aligned
and not worry about the alignment of the stores. Microbenchmarks
confirm that this approach is much faster than the current unaligned
copy loop that uses shifts and rotates to ensure both loads and
stores are aligned.
We also want to try and do the stores in cacheline aligned, cacheline
sized chunks. If the store queue is unable to merge an entire
cacheline of stores then the L2 cache will have to do a
read/modify/write. Even worse, we will serialise this with the stores
in the next iteration of the copy loop since both iterations hit
the same cacheline.
Based on this, the new loop does the following things:
1 - 127 bytes
Get the source 8 byte aligned and use 8 byte loads and stores. Pretty
boring and similar to how the current loop works.
128 - 4095 bytes
Get the source 8 byte aligned and use 8 byte loads and stores,
1 cacheline at a time. We aren't doing the stores in cacheline
aligned chunks so we will potentially serialise once per cacheline.
Even so it is much better than the loop we have today.
4096 - bytes
If both source and destination have the same alignment get them both
16 byte aligned, then get the destination cacheline aligned. Do
cacheline sized loads and stores using VMX.
If source and destination do not have the same alignment, we get the
destination cacheline aligned, and use permute to do aligned loads.
In both cases the VMX loop should be optimal - we always do aligned
loads and stores and are always doing stores in cacheline aligned,
cacheline sized chunks.
To be able to use VMX we must be careful about interrupts and
sleeping. We don't use the VMX loop when in an interrupt (which should
be rare anyway) and we wrap the VMX loop in disable/enable_pagefault
and fall back to the existing copy_tofrom_user loop if we do need to
sleep.
The VMX breakpoint of 4096 bytes was chosen using this microbenchmark:
http://ozlabs.org/~anton/junkcode/copy_to_user.c
Since we are using VMX and there is a cost to saving and restoring
the user VMX state there are two broad cases we need to benchmark:
- Best case - userspace never uses VMX
- Worst case - userspace always uses VMX
In reality a userspace process will sit somewhere between these two
extremes. Since we need to test both aligned and unaligned copies we
end up with 4 combinations. The point at which the VMX loop begins to
win is:
0% VMX
aligned 2048 bytes
unaligned 2048 bytes
100% VMX
aligned 16384 bytes
unaligned 8192 bytes
Considering this is a microbenchmark, the data is hot in cache and
the VMX loop has better store queue merging properties we set the
breakpoint to 4096 bytes, a little below the unaligned breakpoints.
Some future optimisations we can look at:
- Looking at the perf data, a significant part of the cost when a
task is always using VMX is the extra exception we take to restore
the VMX state. As such we should do something similar to the x86
optimisation that restores FPU state for heavy users. ie:
/*
* If the task has used fpu the last 5 timeslices, just do a full
* restore of the math state immediately to avoid the trap; the
* chances of needing FPU soon are obviously high now
*/
preload_fpu = tsk_used_math(next_p) && next_p->fpu_counter > 5;
and
/*
* fpu_counter contains the number of consecutive context switches
* that the FPU is used. If this is over a threshold, the lazy fpu
* saving becomes unlazy to save the trap. This is an unsigned char
* so that after 256 times the counter wraps and the behavior turns
* lazy again; this to deal with bursty apps that only use FPU for
* a short time
*/
- We could create a paca bit to mirror the VMX enabled MSR bit and check
that first, avoiding multiple calls to calling enable_kernel_altivec.
That should help with iovec based system calls like readv.
- We could have two VMX breakpoints, one for when we know the user VMX
state is loaded into the registers and one when it isn't. This could
be a second bit in the paca so we can calculate the break points quickly.
- One suggestion from Ben was to save and restore the VSX registers
we use inline instead of using enable_kernel_altivec.
[BenH: Fixed a problem with preempt and fixed build without CONFIG_ALTIVEC]
Signed-off-by: Anton Blanchard <anton@samba.org>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
kdump fails because we try to execute an HV only instruction. Feature
fixups are being applied after we copy the exception vectors down to 0
so they miss out on any updates.
We have always had this issue but it only became critical in v3.0
when we added CFAR support (breaks POWER5) and v3.1 when we added
POWERNV (breaks everyone).
Signed-off-by: Anton Blanchard <anton@samba.org>
Cc: <stable@kernel.org> [v3.0+]
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
All these files were including module.h just for the basic
EXPORT_SYMBOL infrastructure. We can shift them off to the
export.h header which is a way smaller footprint and thus
realize some compile time gains.
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
* 'merge' of git://git.kernel.org/pub/scm/linux/kernel/git/benh/powerpc: (152 commits)
powerpc: Fix hard CPU IDs detection
powerpc/pmac: Update via-pmu to new syscore_ops
powerpc/kvm: Fix the build for 32-bit Book 3S (classic) processors
powerpc/kvm: Fix kvmppc_core_pending_dec
powerpc: Remove last piece of GEMINI
powerpc: Fix for Pegasos keyboard and mouse
powerpc: Make early memory scan more resilient to out of order nodes
powerpc/pseries/iommu: Cleanup ddw naming
powerpc/pseries/iommu: Find windows after kexec during boot
powerpc/pseries/iommu: Remove ddw property when destroying window
powerpc/pseries/iommu: Add additional checks when changing iommu mask
powerpc/pseries/iommu: Use correct return type in dupe_ddw_if_already_created
powerpc: Remove unused/obsolete CONFIG_XICS
misc: Add CARMA DATA-FPGA Programmer support
misc: Add CARMA DATA-FPGA Access Driver
powerpc: Make IRQ_NOREQUEST last to clear, first to set
powerpc: Integrated Flash controller device tree bindings
powerpc/85xx: Create dts of each core in CAMP mode for P1020RDB
powerpc/85xx: Fix PCIe IDSEL for Px020RDB
powerpc/85xx: P2020 DTS: re-organize dts files
...
Commit e66eed651f ("list: remove prefetching from regular list
iterators") removed the include of prefetch.h from list.h, which
uncovered several cases that had apparently relied on that rather
obscure header file dependency.
So this fixes things up a bit, using
grep -L linux/prefetch.h $(git grep -l '[^a-z_]prefetchw*(' -- '*.[ch]')
grep -L 'prefetchw*(' $(git grep -l 'linux/prefetch.h' -- '*.[ch]')
to guide us in finding files that either need <linux/prefetch.h>
inclusion, or have it despite not needing it.
There are more of them around (mostly network drivers), but this gets
many core ones.
Reported-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Replace all remaining callers of alloc_maybe_bootmem with
zalloc_maybe_bootmem. The callsite in pci_dn is followed with a
memset to clear the memory, and not zeroing at the other callsites
in the celleb fake pci code could lead to following uninitialized
memory as pointers or even freeing said pointers on error paths.
Signed-off-by: Milton Miller <miltonm@bga.com>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
We have a confusing number of ioremap functions. Make things just a
bit simpler by merging ioremap_flags and ioremap_prot.
Signed-off-by: Anton Blanchard <anton@samba.org>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
To make it easier to add optimised versions of copy_page, remove
the 4kB loop for 64kB pages and just do all the work in copy_page.
Signed-off-by: Anton Blanchard <anton@samba.org>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
We check MSR_SF a lot in sstep.c, to decide if we need to emulate the
truncation of values when running in 32-bit mode. Factor out that code
into a helper, and convert it and the other uses to use MSR_64BIT.
This fixes a bug on BOOK3E where kprobes would end up returning to a
32-bit address, because regs->nip was truncated, because (msr & MSR_SF)
was false.
Signed-off-by: Michael Ellerman <michael@ellerman.id.au>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
When we create an alternative feature section, the else case must be the
same size or smaller than the body. This is because when we patch the
else case in we just overwrite the body, so there must be room.
Up to now we just did this by inspection, but it's quite easy to enforce
it in the assembler, so we should.
The only change is to add the ifgt block, but that effects the alignment
of the tabs and so the whole macro is modified.
Also add a test, but #if 0 it because we don't want to break the build.
Anyone who's modifying the feature macros should enable the test.
Signed-off-by: Michael Ellerman <michael@ellerman.id.au>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
The popcnt instructions went into binutils relatively recently. As with a
number of other instructions, create macros and hardcode them.
Signed-off-by: Anton Blanchard <anton@samba.org>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
POWER5 added popcntb, and POWER7 added popcntw and popcntd. As a first step
this patch does all the work out of line, but it would be nice to implement
them as inlines with an out of line fallback.
The performance issue with hweight was noticed when disabling SMT on a large
(192 thread) POWER7 box. The patch improves that testcase by about 8%.
Signed-off-by: Anton Blanchard <anton@samba.org>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Replace EXTRA_CFLAGS with ccflags-y and EXTRA_AFLAGS with asflags-y.
Signed-off-by: matt mooney <mfm@muteddisk.com>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Replace the BOOK3S_64 specific mtmsrd with the generic MTMSRD macro.
Only enable ldstfp when CONFIG_PPC_FPU is set.
Signed-off-by: Sean MacLennan <smaclennan@pikatech.com>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Currently we have the lppaca structs as a simple array of NR_CPUS
entries, taking up space in the data section of the kernel image.
In future we would like to allocate them dynamically, so this
abstracts out the accesses to the array, making it easier to
change how we locate the lppaca for a given cpu in future.
Specifically, lppaca[cpu] changes to lppaca_of(cpu).
Signed-off-by: Paul Mackerras <paulus@samba.org>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
This adds the equivalent of csum_and_copy_from_user for the receive side so we
can copy and checksum in one pass. It is modelled on the generic checksum
routine.
Signed-off-by: Anton Blanchard <anton@samba.org>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
We use the same core loop as the new csum_partial, adding in the
stores and exception handling code. To keep things simple we do all the
exception fixup in csum_and_copy_from_user. This wrapper function is
modelled on the generic checksum code and is careful to always calculate
a complete checksum even if we only copied part of the data to userspace.
To test this I forced checksumming on over loopback and ran socklib (a
simple TCP benchmark). On a POWER6 575 throughput improved by 19% with
this patch. If I forced both the sender and receiver onto the same cpu
(with the hope of shifting the benchmark from being cache bandwidth limited
to cpu limited), adding this patch improved performance by 55%
Signed-off-by: Anton Blanchard <anton@samba.org>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
The main loop of csum_partial runs very slowly on recent POWER CPUs. After some
analysis on both POWER6 and POWER7 I came up with routine below. First we get
the source aligned to a double word, ignoring any odd alignment to keep things
simple. Then we do 64 bytes at a time, with an entry and exit limb of a further
64 bytes. On both POWER6 and POWER7 this should be as fast as we can go since
we are limited by the latency of the adde instructions.
To test this I forced checksumming on over loopback and ran socklib (a
simple TCP benchmark). On a POWER6 575 throughput improved by 11% with
this patch.
Signed-off-by: Anton Blanchard <anton@samba.org>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
The feature-fixup test declare some extern void variables and then take
their addresses. Fix this by declaring them as extern u8 instead.
Fixes these warnings (treated as errors):
CC arch/powerpc/lib/feature-fixups.o
cc1: warnings being treated as errors
arch/powerpc/lib/feature-fixups.c: In function 'test_cpu_macros':
arch/powerpc/lib/feature-fixups.c:293:23: error: taking address of expression of type 'void'
arch/powerpc/lib/feature-fixups.c:294:9: error: taking address of expression of type 'void'
arch/powerpc/lib/feature-fixups.c:297:2: error: taking address of expression of type 'void'
arch/powerpc/lib/feature-fixups.c:297:2: error: taking address of expression of type 'void'
arch/powerpc/lib/feature-fixups.c: In function 'test_fw_macros':
arch/powerpc/lib/feature-fixups.c:306:23: error: taking address of expression of type 'void'
arch/powerpc/lib/feature-fixups.c:307:9: error: taking address of expression of type 'void'
arch/powerpc/lib/feature-fixups.c:310:2: error: taking address of expression of type 'void'
arch/powerpc/lib/feature-fixups.c:310:2: error: taking address of expression of type 'void'
arch/powerpc/lib/feature-fixups.c: In function 'test_lwsync_macros':
arch/powerpc/lib/feature-fixups.c:321:23: error: taking address of expression of type 'void'
arch/powerpc/lib/feature-fixups.c:322:9: error: taking address of expression of type 'void'
arch/powerpc/lib/feature-fixups.c:326:3: error: taking address of expression of type 'void'
arch/powerpc/lib/feature-fixups.c:326:3: error: taking address of expression of type 'void'
arch/powerpc/lib/feature-fixups.c:329:3: error: taking address of expression of type 'void'
arch/powerpc/lib/feature-fixups.c:329:3: error: taking address of expression of type 'void'
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
