When trying to disable the block we bitwise or the control
register with value zero. This is confusing as using bitwise or with
value zero doesn't have any effect at all. Drop this as we already set
the enable bit to zero by appling inverted RNG_RBGEN_MASK.
Signed-off-by: Matthias Brugger <mbrugger@suse.com>
Acked-by: Scott Branden <scott.branden@broadcom.com>
Acked-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>

Counter mode is a stream cipher chaining mode that is typically used
with inputs that are of arbitrarily length, and so a tail block which
is smaller than a full AES block is rule rather than exception.

The current ctr(aes) implementation for arm64 always makes a separate
call into the assembler routine to process this tail block, which is
suboptimal, given that it requires reloading of the AES round keys,
and prevents us from handling this tail block using the 5-way stride
that we use for better performance on deep pipelines.

So let's update the assembler routine so it can handle any input size,
and uses NEON permutation instructions and overlapping loads and stores
to handle the tail block. This results in a ~16% speedup for 1420 byte
blocks on cores with deep pipelines such as ThunderX2.
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>

Commit 69b6f2e8 ("crypto: arm64/aes-neon - limit exposed routines if
faster driver is enabled") intended to hide modes from the plain NEON
driver that are also implemented by the faster bit sliced NEON one if
both are enabled. However, the defined() CPP function does not detect
if the bit sliced NEON driver is enabled as a module. So instead, let's
use IS_ENABLED() here.

Fixes: 69b6f2e8 ("crypto: arm64/aes-neon - limit exposed routines if ...")
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>

Add maintainers for the Intel Keem Bay Offload Crypto Subsystem (OCS)
Hash Control Unit (HCU) crypto driver.
Signed-off-by: Daniele Alessandrelli <daniele.alessandrelli@intel.com>
Acked-by: Declan Murphy <declan.murphy@intel.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>

Add optional support of sha224 and hmac(sha224).
Co-developed-by: Declan Murphy <declan.murphy@intel.com>
Signed-off-by: Declan Murphy <declan.murphy@intel.com>
Signed-off-by: Daniele Alessandrelli <daniele.alessandrelli@intel.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>

Add HMAC support to the Keem Bay OCS HCU driver, thus making it provide
the following additional transformations:
- hmac(sha256)
- hmac(sha384)
- hmac(sha512)
- hmac(sm3)

The Keem Bay OCS HCU hardware does not allow "context-switch" for HMAC
operations, i.e., it does not support computing a partial HMAC, save its
state and then continue it later. Therefore, full hardware acceleration
is provided only when possible (e.g., when crypto_ahash_digest() is
called); in all other cases hardware acceleration is only partial (OPAD
and IPAD calculation is done in software, while hashing is hardware
accelerated).
Co-developed-by: Declan Murphy <declan.murphy@intel.com>
Signed-off-by: Declan Murphy <declan.murphy@intel.com>
Signed-off-by: Daniele Alessandrelli <daniele.alessandrelli@intel.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>

Add support for the Hashing Control Unit (HCU) included in the Offload
Crypto Subsystem (OCS) of the Intel Keem Bay SoC, thus enabling
hardware-accelerated hashing on the Keem Bay SoC for the following
algorithms:
- sha256
- sha384
- sha512
- sm3

The driver is composed of two files:

- 'ocs-hcu.c' which interacts with the hardware and abstracts it by
  providing an API following the usual paradigm used in hashing drivers
  / libraries (e.g., hash_init(), hash_update(), hash_final(), etc.).
  NOTE: this API can block and sleep, since completions are used to wait
  for the HW to complete the hashing.

- 'keembay-ocs-hcu-core.c' which exports the functionality provided by
  'ocs-hcu.c' as a ahash crypto driver. The crypto engine is used to
  provide asynchronous behavior. 'keembay-ocs-hcu-core.c' also takes
  care of the DMA mapping of the input sg list.

The driver passes crypto manager self-tests, including the extra tests
(CRYPTO_MANAGER_EXTRA_TESTS=y).
Signed-off-by: Declan Murphy <declan.murphy@intel.com>
Co-developed-by: Daniele Alessandrelli <daniele.alessandrelli@intel.com>
Signed-off-by: Daniele Alessandrelli <daniele.alessandrelli@intel.com>
Acked-by: Mark Gross <mgross@linux.intel.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>

Add device-tree bindings for the Intel Keem Bay Offload Crypto Subsystem
(OCS) Hashing Control Unit (HCU) crypto driver.
Signed-off-by: Declan Murphy <declan.murphy@intel.com>
Signed-off-by: Daniele Alessandrelli <daniele.alessandrelli@intel.com>
Acked-by: Mark Gross <mgross@linux.intel.com>
Reviewed-by: Rob Herring <robh@kernel.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>

This patchs fixes some remaining style issue.
Signed-off-by: Corentin Labbe <clabbe@baylibre.com>
Signed-off-by: Corentin Labbe <clabbe@baylibre.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>

This patch enable to access usage stats for each algorithm.
Signed-off-by: Corentin Labbe <clabbe@baylibre.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>

With the recent kmap change, some tests which were conditional on
CONFIG_DEBUG_HIGHMEM now are enabled by default.
This permit to detect a problem in sun4i-ss usage of kmap.

sun4i-ss uses two kmap via sg_miter (one for input, one for output), but
using two kmap at the same time is hard:
"the ordering has to be correct and with sg_miter that's probably hard to get
right." (quoting Tlgx)

So the easiest solution is to never have two sg_miter/kmap open at the same time.
After each use of sg_miter, I store the current index, for being able to
resume sg_miter to the right place.

Fixes: 6298e948 ("crypto: sunxi-ss - Add Allwinner Security System crypto accelerator")
Signed-off-by: Corentin Labbe <clabbe@baylibre.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>

The need_fallback is never initialized and seem to be always true at runtime.
So all hardware operations are always bypassed.

Fixes: 0ae1f46c ("crypto: sun4i-ss - fallback when length is not multiple of blocksize")
Cc: <stable@vger.kernel.org>
Signed-off-by: Corentin Labbe <clabbe@baylibre.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>

Ciphers produce invalid results on BE.
Key and IV need to be written in LE.

Fixes: 6298e948 ("crypto: sunxi-ss - Add Allwinner Security System crypto accelerator")
Cc: <stable@vger.kernel.org>
Signed-off-by: Corentin Labbe <clabbe@baylibre.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>

Allwinner A10 and A13 SoC have a version of the SS which produce
invalid IV in IVx register.

Instead of adding a variant for those, let's convert SS to produce IV
directly from data.
Fixes: 6298e948 ("crypto: sunxi-ss - Add Allwinner Security System crypto accelerator")
Cc: <stable@vger.kernel.org>
Signed-off-by: Corentin Labbe <clabbe@baylibre.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>

The optimized cipher function need length multiple of 4 bytes.
But it get sometimes odd length.
This is due to SG data could be stored with an offset.

So the fix is to check also if the offset is aligned with 4 bytes.
Fixes: 6298e948 ("crypto: sunxi-ss - Add Allwinner Security System crypto accelerator")
Cc: <stable@vger.kernel.org>
Signed-off-by: Corentin Labbe <clabbe@baylibre.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>

When running the non-optimized cipher function, SS produce partial random
output.
This is due to linearize buffers being reseted after each loop.

For preserving stack, instead of moving them back to start of function,
I move them in sun4i_ss_ctx.

Fixes: 8d3bcb99 ("crypto: sun4i-ss - reduce stack usage")
Signed-off-by: Corentin Labbe <clabbe@baylibre.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>

Remove dev_err() messages after platform_get_irq*() failures.
drivers/crypto/inside-secure/safexcel.c: line 1161 is redundant
because platform_get_irq() already prints an error

Generated by: scripts/coccinelle/api/platform_get_irq.cocci
Signed-off-by: Tian Tao <tiantao6@hisilicon.com>
Acked-by: Antoine Tenart <atenart@kernel.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>

The cipher routines in the crypto API are mostly intended for templates
implementing skcipher modes generically in software, and shouldn't be
used outside of the crypto subsystem. So move the prototypes and all
related definitions to a new header file under include/crypto/internal.
Also, let's use the new module namespace feature to move the symbol
exports into a new namespace CRYPTO_INTERNAL.
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Acked-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>

Allocating a cipher via the crypto API only to free it again after using
it to encrypt a single block is unnecessary in cases where the algorithm
is known at compile time. So replace this pattern with a call to the AES
library.

Cc: Ayush Sawal <ayush.sawal@chelsio.com>
Cc: Vinay Kumar Yadav <vinay.yadav@chelsio.com>
Cc: Rohit Maheshwari <rohitm@chelsio.com>
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Reviewed-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>

Remove including <linux/version.h> that don't need it.
Signed-off-by: Tian Tao <tiantao6@hisilicon.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>

Since 5.10-rc1 i.MX is a devicetree-only platform and the existing
.id_table support in this driver was only useful for old non-devicetree
platforms.

Remove the unused .id_table support.
Signed-off-by: Fabio Estevam <festevam@gmail.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>

The signed long type used for printing the number of bytes processed in
tcrypt benchmarks limits the range to -/+ 2 GiB, which is not sufficient
to cover the performance of common accelerated ciphers such as AES-NI
when benchmarked with sec=1. So switch to u64 instead.

While at it, fix up a missing printk->pr_cont conversion in the AEAD
benchmark.
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>

Follow the same approach as the arm64 driver for implementing a version
of AES-NI in CBC mode that supports ciphertext stealing. This results in
a ~2x speed increase for relatively short inputs (less than 256 bytes),
which is relevant given that AES-CBC with ciphertext stealing is used
for filename encryption in the fscrypt layer. For larger inputs, the
speedup is still significant (~25% on decryption, ~6% on encryption)

Tested-by: Eric Biggers <ebiggers@google.com> # x86_64
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>

E-mails to Kamil Konieczny to his Samsung address bounce with 550 (User
unknown).  Kamil no longer takes care about Samsung S5P SSS driver so
remove the invalid email address from:
 - mailmap,
 - bindings maintainer entries,
 - maintainers entry for S5P Security Subsystem crypto accelerator.
Signed-off-by: Krzysztof Kozlowski <krzk@kernel.org>
Acked-by: Vladimir Zapolskiy <vz@mleia.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>

The crypto mediatek driver has been replaced by the inside-secure
driver now. Remove this driver to avoid having duplicate drivers.
Signed-off-by: Vic Wu <vic.wu@mediatek.com>
Acked-by: Ryder Lee <ryder.lee@mediatek.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>

Pavel reports that commit 17858b14 ("crypto: ecdh - avoid unaligned
accesses in ecdh_set_secret()") fixes one problem but introduces another:
the unconditional memcpy() introduced by that commit may overflow the
target buffer if the source data is invalid, which could be the result of
intentional tampering.

So check params.key_size explicitly against the size of the target buffer
before validating the key further.

Fixes: 17858b14 ("crypto: ecdh - avoid unaligned accesses in ecdh_set_secret()")
Reported-by: Pavel Machek <pavel@denx.de>
Cc: <stable@vger.kernel.org>
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>

Commit 86cd97ec ("crypto: arm/chacha-neon - optimize for non-block
size multiples") refactored the chacha block handling in the glue code in
a way that may result in the counter increment to be omitted when calling
chacha_block_xor_neon() to process a full block. This violates the skcipher
API, which requires that the output IV is suitable for handling more input
as long as the preceding input has been presented in round multiples of the
block size. Also, the same code is exposed via the chacha library interface
whose callers may actually rely on this increment to occur even for final
blocks that are smaller than the chacha block size.

So increment the counter after calling chacha_block_xor_neon().

Fixes: 86cd97ec ("crypto: arm/chacha-neon - optimize for non-block size multiples")
Reported-by: Eric Biggers <ebiggers@kernel.org>
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>

Since commit 36e2c742 ("fs: don't allow splice read/write without
explicit ops") we've required that file operation structures explicitly
enable splice support, rather than falling back to the default handlers.

Most /proc files use the indirect 'struct proc_ops' to describe their
file operations, and were fixed up to support splice earlier in commits
40be821d..b24c30c6, but the mountinfo files interact with the
VFS directly using their own 'struct file_operations' and got missed as
a result.

This adds the necessary support for splice to work for /proc/*/mountinfo
and friends.
Reported-by: Joan Bruguera Micó <joanbrugueram@gmail.com>
Reported-by: Jussi Kivilinna <jussi.kivilinna@iki.fi>
Link: https://bugzilla.kernel.org/show_bug.cgi?id=209971
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Christoph Hellwig <hch@lst.de>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

Pull NTB fixes from Jon Mason:
 "Bug fix for IDT NTB and Intel NTB LTR management support"

* tag 'ntb-5.11' of git://github.com/jonmason/ntb:
  ntb: intel: add Intel NTB LTR vendor support for gen4 NTB
  ntb: idt: fix error check in ntb_hw_idt.c

Pull crypto fixes from Herbert Xu:
 "Fix a number of autobuild failures due to missing Kconfig
  dependencies"

* 'linus' of git://git.kernel.org/pub/scm/linux/kernel/git/herbert/crypto-2.6:
  crypto: qat - add CRYPTO_AES to Kconfig dependencies
  crypto: keembay - Add dependency on HAS_IOMEM
  crypto: keembay - CRYPTO_DEV_KEEMBAY_OCS_AES_SM4 should depend on ARCH_KEEMBAY

Pull objtool fix from Ingo Molnar:
 "Fix a segfault that occurs when built with Clang"

* tag 'objtool-urgent-2020-12-27' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  objtool: Fix seg fault with Clang non-section symbols

Pull locking fixes from Ingo Molnar:
 "Misc fixes/updates:

   - Fix static keys usage in module __init sections

   - Add separate MAINTAINERS entry for static branches/calls

   - Fix lockdep splat with CONFIG_PREEMPTIRQ_EVENTS=y tracing"

* tag 'locking-urgent-2020-12-27' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  softirq: Avoid bad tracing / lockdep interaction
  jump_label/static_call: Add MAINTAINERS
  jump_label: Fix usage in module __init

Pull timer fixes from Ingo Molnar:
 "Update/fix two CPU sanity checks in the hotplug and the boot code, and
  fix a typo in the Kconfig help text.

  [ Context: the first two commits are the result of an ongoing
    annotation+review work of (intentional) tick_do_timer_cpu() data
    races reported by KCSAN, but the annotations aren't fully cooked
    yet ]"

* tag 'timers-urgent-2020-12-27' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  timekeeping: Fix spelling mistake in Kconfig "fullfill" -> "fulfill"
  tick/sched: Remove bogus boot "safety" check
  tick: Remove pointless cpu valid check in hotplug code

Pull scheduler fix from Ingo Molnar:
 "Fix a context switch performance regression"

* tag 'sched-urgent-2020-12-27' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  sched: Optimize finish_lock_switch()

Commit c9a3c4e6 ("mfd: ab8500-debugfs: Remove extraneous curly
brace") removed a left-over curly brace that caused build failures, but
Joe Perches points out that the subsequent 'seq_putc()' should also be
removed, because the commit that caused all these problems already added
the final '\n' to the seq_printf() above it.
Reported-by: Joe Perches <joe@perches.com>
Fixes: 886c8121 ("mfd: ab8500-debugfs: Remove the racy fiddling with irq_desc")
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Nathan Chancellor <natechancellor@gmail.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

Pull PCI fixes from Bjorn Helgaas:

 - Fix a tegra enumeration regression (Rob Herring)

 - Fix a designware-host check that warned on *success*, not failure
   (Alexander Lobakin)

* tag 'pci-v5.11-fixes-1' of git://git.kernel.org/pub/scm/linux/kernel/git/helgaas/pci:
  PCI: dwc: Fix inverted condition of DMA mask setup warning
  PCI: tegra: Fix host link initialization

Clang errors:

  drivers/mfd/ab8500-debugfs.c:1526:2: error: non-void function does not return a value [-Werror,-Wreturn-type]
          }
          ^
  drivers/mfd/ab8500-debugfs.c:1528:2: error: expected identifier or '('
  return 0;
          ^
  drivers/mfd/ab8500-debugfs.c:1529:1: error: extraneous closing brace ('}')
  }
  ^
  3 errors generated.

The cleanup in ab8500_interrupts_show left a curly brace around, remove
it to fix the error.

Fixes: 886c8121 ("mfd: ab8500-debugfs: Remove the racy fiddling with irq_desc")
Signed-off-by: Nathan Chancellor <natechancellor@gmail.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

Commit 660c4865 ("PCI: dwc: Set 32-bit DMA mask for MSI target address
allocation") added dma_mask_set() call to explicitly set 32-bit DMA mask
for MSI message mapping, but for now it throws a warning on ret == 0, while
dma_set_mask() returns 0 in case of success.

Fix this by inverting the condition.

[bhelgaas: join string to make it greppable]
Fixes: 660c4865 ("PCI: dwc: Set 32-bit DMA mask for MSI target address allocation")
Link: https://lore.kernel.org/r/20201222150708.67983-1-alobakin@pm.meSigned-off-by: Alexander Lobakin <alobakin@pm.me>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>

Commit b9ac0f9d ("PCI: dwc: Move dw_pcie_setup_rc() to DWC common
code") broke enumeration of downstream devices on Tegra:

In non-working case (next-20201211):

  0001:00:00.0 PCI bridge: NVIDIA Corporation Device 1ad2 (rev a1)
  0001:01:00.0 SATA controller: Marvell Technology Group Ltd. Device 9171 (rev 13)
  0005:00:00.0 PCI bridge: NVIDIA Corporation Device 1ad0 (rev a1)

In working case (v5.10-rc7):

  0001:00:00.0 PCI bridge: Molex Incorporated Device 1ad2 (rev a1)
  0001:01:00.0 SATA controller: Marvell Technology Group Ltd. Device 9171 (rev 13)
  0005:00:00.0 PCI bridge: Molex Incorporated Device 1ad0 (rev a1)
  0005:01:00.0 PCI bridge: PLX Technology, Inc. Device 3380 (rev ab)
  0005:02:02.0 PCI bridge: PLX Technology, Inc. Device 3380 (rev ab)
  0005:03:00.0 USB controller: PLX Technology, Inc. Device 3380 (rev ab)

The problem seems to be dw_pcie_setup_rc() is now called twice before and
after the link up handling. The fix is to move Tegra's link up handling to
.start_link() function like other DWC drivers. Tegra is a bit more
complicated than others as it re-inits the whole DWC controller to retry
the link. With this, the initialization ordering is restored to match the
prior sequence.

Fixes: b9ac0f9d ("PCI: dwc: Move dw_pcie_setup_rc() to DWC common code")
Link: https://lore.kernel.org/r/20201218143905.1614098-1-robh@kernel.orgReported-by: Mian Yousaf Kaukab <ykaukab@suse.de>
Tested-by: Mian Yousaf Kaukab <ykaukab@suse.de>
Signed-off-by: Rob Herring <robh@kernel.org>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Cc: Lorenzo Pieralisi <lorenzo.pieralisi@arm.com>
Cc: Thierry Reding <thierry.reding@gmail.com>
Cc: Jonathan Hunter <jonathanh@nvidia.com>
Cc: Vidya Sagar <vidyas@nvidia.com>