Skip to content

K
Kernel

openeuler / Kernel
12 个月 前同步成功

通知 5
Star 0
Fork 0
K
Kernel

体验新版 GitCode，发现更多精彩内容 >>

切换分支/标签
  1. 11 10月, 2007 3 次提交
  3. 13 7月, 2007 2 次提交
    • D
      async_tx: add the async_tx api · 9bc89cd8
      由 Dan Williams 提交于 
      The async_tx api provides methods for describing a chain of asynchronous
bulk memory transfers/transforms with support for inter-transactional
dependencies.  It is implemented as a dmaengine client that smooths over
the details of different hardware offload engine implementations.  Code
that is written to the api can optimize for asynchronous operation and the
api will fit the chain of operations to the available offload resources. 
 
	I imagine that any piece of ADMA hardware would register with the
	'async_*' subsystem, and a call to async_X would be routed as
	appropriate, or be run in-line. - Neil Brown

async_tx exploits the capabilities of struct dma_async_tx_descriptor to
provide an api of the following general format:

struct dma_async_tx_descriptor *
async_<operation>(..., struct dma_async_tx_descriptor *depend_tx,
			dma_async_tx_callback cb_fn, void *cb_param)
{
	struct dma_chan *chan = async_tx_find_channel(depend_tx, <operation>);
	struct dma_device *device = chan ? chan->device : NULL;
	int int_en = cb_fn ? 1 : 0;
	struct dma_async_tx_descriptor *tx = device ?
		device->device_prep_dma_<operation>(chan, len, int_en) : NULL;

	if (tx) { /* run <operation> asynchronously */
		...
		tx->tx_set_dest(addr, tx, index);
		...
		tx->tx_set_src(addr, tx, index);
		...
		async_tx_submit(chan, tx, flags, depend_tx, cb_fn, cb_param);
	} else { /* run <operation> synchronously */
		...
		<operation>
		...
		async_tx_sync_epilog(flags, depend_tx, cb_fn, cb_param);
	}

	return tx;
}

async_tx_find_channel() returns a capable channel from its pool.  The
channel pool is organized as a per-cpu array of channel pointers.  The
async_tx_rebalance() routine is tasked with managing these arrays.  In the
uniprocessor case async_tx_rebalance() tries to spread responsibility
evenly over channels of similar capabilities.  For example if there are two
copy+xor channels, one will handle copy operations and the other will
handle xor.  In the SMP case async_tx_rebalance() attempts to spread the
operations evenly over the cpus, e.g. cpu0 gets copy channel0 and xor
channel0 while cpu1 gets copy channel 1 and xor channel 1.  When a
dependency is specified async_tx_find_channel defaults to keeping the
operation on the same channel.  A xor->copy->xor chain will stay on one
channel if it supports both operation types, otherwise the transaction will
transition between a copy and a xor resource.

Currently the raid5 implementation in the MD raid456 driver has been
converted to the async_tx api.  A driver for the offload engines on the
Intel Xscale series of I/O processors, iop-adma, is provided in a later
commit.  With the iop-adma driver and async_tx, raid456 is able to offload
copy, xor, and xor-zero-sum operations to hardware engines.
 
On iop342 tiobench showed higher throughput for sequential writes (20 - 30%
improvement) and sequential reads to a degraded array (40 - 55%
improvement).  For the other cases performance was roughly equal, +/- a few
percentage points.  On a x86-smp platform the performance of the async_tx
implementation (in synchronous mode) was also +/- a few percentage points
of the original implementation.  According to 'top' on iop342 CPU
utilization drops from ~50% to ~15% during a 'resync' while the speed
according to /proc/mdstat doubles from ~25 MB/s to ~50 MB/s.
 
The tiobench command line used for testing was: tiobench --size 2048
--block 4096 --block 131072 --dir /mnt/raid --numruns 5
* iop342 had 1GB of memory available

Details:
* if CONFIG_DMA_ENGINE=n the asynchronous path is compiled away by making
  async_tx_find_channel a static inline routine that always returns NULL
* when a callback is specified for a given transaction an interrupt will
  fire at operation completion time and the callback will occur in a
  tasklet.  if the the channel does not support interrupts then a live
  polling wait will be performed
* the api is written as a dmaengine client that requests all available
  channels
* In support of dependencies the api implicitly schedules channel-switch
  interrupts.  The interrupt triggers the cleanup tasklet which causes
  pending operations to be scheduled on the next channel
* Xor engines treat an xor destination address differently than a software
  xor routine.  To the software routine the destination address is an implied
  source, whereas engines treat it as a write-only destination.  This patch
  modifies the xor_blocks routine to take a an explicit destination address
  to mirror the hardware.

Changelog:
* fixed a leftover debug print
* don't allow callbacks in async_interrupt_cond
* fixed xor_block changes
* fixed usage of ASYNC_TX_XOR_DROP_DEST
* drop dma mapping methods, suggested by Chris Leech
* printk warning fixups from Andrew Morton
* don't use inline in C files, Adrian Bunk
* select the API when MD is enabled
* BUG_ON xor source counts <= 1
* implicitly handle hardware concerns like channel switching and
  interrupts, Neil Brown
* remove the per operation type list, and distribute operation capabilities
  evenly amongst the available channels
* simplify async_tx_find_channel to optimize the fast path
* introduce the channel_table_initialized flag to prevent early calls to
  the api
* reorganize the code to mimic crypto
* include mm.h as not all archs include it in dma-mapping.h
* make the Kconfig options non-user visible, Adrian Bunk
* move async_tx under crypto since it is meant as 'core' functionality, and
  the two may share algorithms in the future
* move large inline functions into c files
* checkpatch.pl fixes
* gpl v2 only correction

Cc: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: NDan Williams <dan.j.williams@intel.com>
Acked-By: NNeilBrown <neilb@suse.de>
      9bc89cd8
    • D
      xor: make 'xor_blocks' a library routine for use with async_tx · 685784aa
      由 Dan Williams 提交于 
      The async_tx api tries to use a dma engine for an operation, but will fall
back to an optimized software routine otherwise.  Xor support is
implemented using the raid5 xor routines.  For organizational purposes this
routine is moved to a common area.

The following fixes are also made:
* rename xor_block => xor_blocks, suggested by Adrian Bunk
* ensure that xor.o initializes before md.o in the built-in case
* checkpatch.pl fixes
* mark calibrate_xor_blocks __init, Adrian Bunk

Cc: Adrian Bunk <bunk@stusta.de>
Cc: NeilBrown <neilb@suse.de>
Cc: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: NDan Williams <dan.j.williams@intel.com>
      685784aa
  5. 11 7月, 2007 1 次提交
  7. 09 5月, 2007 1 次提交
  9. 02 5月, 2007 2 次提交
  11. 07 2月, 2007 3 次提交
  13. 06 2月, 2007 1 次提交
  15. 07 12月, 2006 3 次提交
  17. 16 10月, 2006 1 次提交
  19. 21 9月, 2006 11 次提交
  21. 26 6月, 2006 1 次提交
  23. 10 1月, 2006 1 次提交
  25. 07 1月, 2006 4 次提交
  27. 02 9月, 2005 1 次提交
  29. 07 7月, 2005 1 次提交
    • A
      [CRYPTO] Add x86_64 asm AES · a2a892a2
      由 Andreas Steinmetz 提交于 
      Implementation:
===============
The encrypt/decrypt code is based on an x86 implementation I did a while
ago which I never published. This unpublished implementation does
include an assembler based key schedule and precomputed tables. For
simplicity and best acceptance, however, I took Gladman's in-kernel code
for table generation and key schedule for the kernel port of my
assembler code and modified this code to produce the key schedule as
required by my assembler implementation. File locations and Kconfig are
kept similar to the i586 AES assembler implementation.
It may seem a little bit strange to use 32 bit I/O and registers in the
assembler implementation but this gives the best code size. My
implementation takes one instruction more per round compared to
Gladman's x86 assembler but it doesn't require any stack for local
variables or saved registers and it is less serialized than Gladman's
code.
Note that all comparisons to Gladman's code were done after my code was
implemented. I did only use FIPS PUB 197 for the implementation so my
implementation is independent work.
If anybody has a better assembler solution for x86_64 I'll be pleased to
have my code replaced with the better solution.

Testing:
========
The implementation passes the in-kernel crypto testing module and I'm
running it without any problems on my laptop where it is mainly used for
dm-crypt.

Microbenchmark:
===============
The microbenchmark was done in userspace with similar compile flags as
used during kernel compile.
Encrypt/decrypt is about 35% faster than the generic C implementation.
As the generic C as well as my assembler implementation are both table
I don't really expect that there is much room for further
improvements though I'll be glad to be corrected here.
The key schedule is about 5% slower than the generic C implementation.
This is due to the fact that some more work has to be done in the key
schedule routine to fit the schedule to the assembler implementation.

Code Size:
==========
Encrypt and decrypt are together about 2.1 Kbytes smaller than the
generic C implementation which is important with regard to L1 cache
usage. The key schedule routine is about 100 bytes larger than the
generic C implementation.

Data Size:
==========
There's no difference in data size requirements between the assembler
implementation and the generic C implementation.

License:
========
Gladmans's code is dual BSD/GPL whereas my assembler code is GPLv2 only
(I'm  not going to change the license for my code). So I had to change
the module license for the x86_64 aes module from 'Dual BSD/GPL' to
'GPL' to reflect the most restrictive license within the module.
Signed-off-by: NAndreas Steinmetz <ast@domdv.de>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>
      a2a892a2
  31. 01 5月, 2005 1 次提交
  33. 17 4月, 2005 1 次提交
    • L
      Linux-2.6.12-rc2 · 1da177e4
      由 Linus Torvalds 提交于 
      Initial git repository build. I'm not bothering with the full history,
even though we have it. We can create a separate "historical" git
archive of that later if we want to, and in the meantime it's about
3.2GB when imported into git - space that would just make the early
git days unnecessarily complicated, when we don't have a lot of good
infrastructure for it.

Let it rip!
      1da177e4