Pull md updates from Neil Brown:
"More updates that usual this time. A few have performance impacts
which hould mostly be positive, but RAID5 (in particular) can be very
work-load ensitive... We'll have to wait and see.
Highlights:
- "experimental" code for managing md/raid1 across a cluster using
DLM. Code is not ready for general use and triggers a WARNING if
used. However it is looking good and mostly done and having in
mainline will help co-ordinate development.
- RAID5/6 can now batch multiple (4K wide) stripe_heads so as to
handle a full (chunk wide) stripe as a single unit.
- RAID6 can now perform read-modify-write cycles which should help
performance on larger arrays: 6 or more devices.
- RAID5/6 stripe cache now grows and shrinks dynamically. The value
set is used as a minimum.
- Resync is now allowed to go a little faster than the 'mininum' when
there is competing IO. How much faster depends on the speed of the
devices, so the effective minimum should scale with device speed to
some extent"
* tag 'md/4.1' of git://neil.brown.name/md: (58 commits)
md/raid5: don't do chunk aligned read on degraded array.
md/raid5: allow the stripe_cache to grow and shrink.
md/raid5: change ->inactive_blocked to a bit-flag.
md/raid5: move max_nr_stripes management into grow_one_stripe and drop_one_stripe
md/raid5: pass gfp_t arg to grow_one_stripe()
md/raid5: introduce configuration option rmw_level
md/raid5: activate raid6 rmw feature
md/raid6 algorithms: xor_syndrome() for SSE2
md/raid6 algorithms: xor_syndrome() for generic int
md/raid6 algorithms: improve test program
md/raid6 algorithms: delta syndrome functions
raid5: handle expansion/resync case with stripe batching
raid5: handle io error of batch list
RAID5: batch adjacent full stripe write
raid5: track overwrite disk count
raid5: add a new flag to track if a stripe can be batched
raid5: use flex_array for scribble data
md raid0: access mddev->queue (request queue member) conditionally because it is not set when accessed from dm-raid
md: allow resync to go faster when there is competing IO.
md: remove 'go_faster' option from ->sync_request()
...
Glue it altogehter. The raid6 rmw path should work the same as the
already existing raid5 logic. So emulate the prexor handling/flags
and split functions as needed.
1) Enable xor_syndrome() in the async layer.
2) Split ops_run_prexor() into RAID4/5 and RAID6 logic. Xor the syndrome
at the start of a rmw run as we did it before for the single parity.
3) Take care of rmw run in ops_run_reconstruct6(). Again process only
the changed pages to get syndrome back into sync.
4) Enhance set_syndrome_sources() to fill NULL pages if we are in a rmw
run. The lower layers will calculate start & end pages from that and
call the xor_syndrome() correspondingly.
5) Adapt the several places where we ignored Q handling up to now.
Performance numbers for a single E5630 system with a mix of 10 7200k
desktop/server disks. 300 seconds random write with 8 threads onto a
3,2TB (10*400GB) RAID6 64K chunk without spare (group_thread_cnt=4)
bsize rmw_level=1 rmw_level=0 rmw_level=1 rmw_level=0
skip_copy=1 skip_copy=1 skip_copy=0 skip_copy=0
4K 115 KB/s 141 KB/s 165 KB/s 140 KB/s
8K 225 KB/s 275 KB/s 324 KB/s 274 KB/s
16K 434 KB/s 536 KB/s 640 KB/s 534 KB/s
32K 751 KB/s 1,051 KB/s 1,234 KB/s 1,045 KB/s
64K 1,339 KB/s 1,958 KB/s 2,282 KB/s 1,962 KB/s
128K 2,673 KB/s 3,862 KB/s 4,113 KB/s 3,898 KB/s
256K 7,685 KB/s 7,539 KB/s 7,557 KB/s 7,638 KB/s
512K 19,556 KB/s 19,558 KB/s 19,652 KB/s 19,688 Kb/s
Signed-off-by: Markus Stockhausen <stockhausen@collogia.de>
Signed-off-by: NeilBrown <neilb@suse.de>
Commit 9c521a200b ("crypto: api - remove instance when test failed")
tried to grab a module reference count before the module was even set.
Worse, it then goes on to free the module reference count after it is
set so you quickly end up with a negative module reference count which
prevents people from using any instances belonging to that module.
This patch moves the module initialisation before the reference
count.
Reported-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The networking updates from David Miller removed the iocb argument from
sendmsg and recvmsg (in commit 1b78414047: "net: Remove iocb argument
from sendmsg and recvmsg"), but the crypto code had added new instances
of them.
When I pulled the crypto update, it was a silent semantic mis-merge, and
I overlooked the new warning messages in my test-build. I try to fix
those in the merge itself, but that relies on me noticing. Oh well.
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Pull crypto update from Herbert Xu:
"Here is the crypto update for 4.1:
New interfaces:
- user-space interface for AEAD
- user-space interface for RNG (i.e., pseudo RNG)
New hashes:
- ARMv8 SHA1/256
- ARMv8 AES
- ARMv8 GHASH
- ARM assembler and NEON SHA256
- MIPS OCTEON SHA1/256/512
- MIPS img-hash SHA1/256 and MD5
- Power 8 VMX AES/CBC/CTR/GHASH
- PPC assembler AES, SHA1/256 and MD5
- Broadcom IPROC RNG driver
Cleanups/fixes:
- prevent internal helper algos from being exposed to user-space
- merge common code from assembly/C SHA implementations
- misc fixes"
* git://git.kernel.org/pub/scm/linux/kernel/git/herbert/crypto-2.6: (169 commits)
crypto: arm - workaround for building with old binutils
crypto: arm/sha256 - avoid sha256 code on ARMv7-M
crypto: x86/sha512_ssse3 - move SHA-384/512 SSSE3 implementation to base layer
crypto: x86/sha256_ssse3 - move SHA-224/256 SSSE3 implementation to base layer
crypto: x86/sha1_ssse3 - move SHA-1 SSSE3 implementation to base layer
crypto: arm64/sha2-ce - move SHA-224/256 ARMv8 implementation to base layer
crypto: arm64/sha1-ce - move SHA-1 ARMv8 implementation to base layer
crypto: arm/sha2-ce - move SHA-224/256 ARMv8 implementation to base layer
crypto: arm/sha256 - move SHA-224/256 ASM/NEON implementation to base layer
crypto: arm/sha1-ce - move SHA-1 ARMv8 implementation to base layer
crypto: arm/sha1_neon - move SHA-1 NEON implementation to base layer
crypto: arm/sha1 - move SHA-1 ARM asm implementation to base layer
crypto: sha512-generic - move to generic glue implementation
crypto: sha256-generic - move to generic glue implementation
crypto: sha1-generic - move to generic glue implementation
crypto: sha512 - implement base layer for SHA-512
crypto: sha256 - implement base layer for SHA-256
crypto: sha1 - implement base layer for SHA-1
crypto: api - remove instance when test failed
crypto: api - Move alg ref count init to crypto_check_alg
...
This updated the generic SHA-512 implementation to use the
generic shared SHA-512 glue code.
It also implements a .finup hook crypto_sha512_finup() and exports
it to other modules. The import and export() functions and the
.statesize member are dropped, since the default implementation
is perfectly suitable for this module.
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
This updates the generic SHA-256 implementation to use the
new shared SHA-256 glue code.
It also implements a .finup hook crypto_sha256_finup() and exports
it to other modules. The import and export() functions and the
.statesize member are dropped, since the default implementation
is perfectly suitable for this module.
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
This updated the generic SHA-1 implementation to use the generic
shared SHA-1 glue code.
It also implements a .finup hook crypto_sha1_finup() and exports
it to other modules. The import and export() functions and the
.statesize member are dropped, since the default implementation
is perfectly suitable for this module.
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
A cipher instance is added to the list of instances unconditionally
regardless of whether the associated test failed. However, a failed
test implies that during another lookup, the cipher instance will
be added to the list again as it will not be found by the lookup
code.
That means that the list can be filled up with instances whose tests
failed.
Note: tests only fail in reality in FIPS mode when a cipher is not
marked as fips_allowed=1. This can be seen with cmac(des3_ede) that does
not have a fips_allowed=1. When allocating the cipher, the allocation
fails with -ENOENT due to the missing fips_allowed=1 flag (which
causes the testmgr to return EINVAL). Yet, the instance of
cmac(des3_ede) is shown in /proc/crypto. Allocating the cipher again
fails again, but a 2nd instance is listed in /proc/crypto.
The patch simply de-registers the instance when the testing failed.
Signed-off-by: Stephan Mueller <smueller@chronox.de>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
We currently initialise the crypto_alg ref count in the function
__crypto_register_alg. As one of the callers of that function
crypto_register_instance needs to obtain a ref count before it
calls __crypto_register_alg, we need to move the initialisation
out of there.
Since both callers of __crypto_register_alg call crypto_check_alg,
this is the logical place to perform the initialisation.
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Acked-by: Stephan Mueller <smueller@chronox.de>
trivial conflict in net/socket.c and non-trivial one in crypto -
that one had evaded aio_complete() removal.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
The function crypto_alg_match returns an algorithm without taking
any references on it. This means that the algorithm can be freed
at any time, therefore all users of crypto_alg_match are buggy.
This patch fixes this by taking a reference count on the algorithm
to prevent such races.
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
This patch makes crypto_unregister_instance take a crypto_instance
instead of a crypto_alg. This allows us to remove a duplicate
CRYPTO_ALG_INSTANCE check in crypto_unregister_instance.
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
There are multiple problems in crypto_unregister_instance:
1) The cra_refcnt BUG_ON check is racy and can cause crashes.
2) The cra_refcnt check shouldn't exist at all.
3) There is no reference on tmpl to protect the tmpl->free call.
This patch rewrites the function using crypto_remove_spawn which
now morphs into crypto_remove_instance.
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
After the TX sgl is expanded we need to explicitly mark end of data
at the last buffer that contains data.
Changes in v2
- use type 'bool' and true/false for 'mark'.
Signed-off-by: Tadeusz Struk <tadeusz.struk@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
No need to use kzalloc to allocate sgls as the structure is initialized anyway.
Signed-off-by: Tadeusz Struk <tadeusz.struk@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The mcryptd is used as a wrapper around internal ciphers. Therefore,
the mcryptd must process the internal cipher by marking mcryptd as
internal if the underlying cipher is an internal cipher.
Signed-off-by: Stephan Mueller <smueller@chronox.de>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
With ciphers that now cannot be accessed via the kernel crypto API,
callers shall be able to identify the ciphers that are not callable. The
/proc/crypto file is added a boolean field identifying that such
internal ciphers.
Signed-off-by: Stephan Mueller <smueller@chronox.de>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
The cryptd is used as a wrapper around internal ciphers. Therefore, the
cryptd must process the internal cipher by marking cryptd as internal if
the underlying cipher is an internal cipher.
Signed-off-by: Stephan Mueller <smueller@chronox.de>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Allocate the ciphers irrespectively if they are marked as internal
or not. As all ciphers, including the internal ciphers will be
processed by the testmgr, it needs to be able to allocate those
ciphers.
Signed-off-by: Stephan Mueller <smueller@chronox.de>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Several hardware related cipher implementations are implemented as
follows: a "helper" cipher implementation is registered with the
kernel crypto API.
Such helper ciphers are never intended to be called by normal users. In
some cases, calling them via the normal crypto API may even cause
failures including kernel crashes. In a normal case, the "wrapping"
ciphers that use the helpers ensure that these helpers are invoked
such that they cannot cause any calamity.
Considering the AF_ALG user space interface, unprivileged users can
call all ciphers registered with the crypto API, including these
helper ciphers that are not intended to be called directly. That
means, with AF_ALG user space may invoke these helper ciphers
and may cause undefined states or side effects.
To avoid any potential side effects with such helpers, the patch
prevents the helpers to be called directly. A new cipher type
flag is added: CRYPTO_ALG_INTERNAL. This flag shall be used
to mark helper ciphers. These ciphers can only be used if the
caller invoke the cipher with CRYPTO_ALG_INTERNAL in the type and
mask field.
Signed-off-by: Stephan Mueller <smueller@chronox.de>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Change type from unsigned long to int to fix an issue reported by kbuild robot:
crypto/algif_skcipher.c:596 skcipher_recvmsg_async() warn: unsigned 'used' is
never less than zero.
Signed-off-by: Tadeusz Struk <tadeusz.struk@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
