Merge tag 'v6.16-p1' of git://git.kernel.org/pub/scm/linux/kernel/git/herbert/crypto-2.6

Pull crypto updates from Herbert Xu:
 "API:
   - Fix memcpy_sglist to handle partially overlapping SG lists
   - Use memcpy_sglist to replace null skcipher
   - Rename CRYPTO_TESTS to CRYPTO_BENCHMARK
   - Flip CRYPTO_MANAGER_DISABLE_TEST into CRYPTO_SELFTESTS
   - Hide CRYPTO_MANAGER
   - Add delayed freeing of driver crypto_alg structures

  Compression:
   - Allocate large buffers on first use instead of initialisation in scomp
   - Drop destination linearisation buffer in scomp
   - Move scomp stream allocation into acomp
   - Add acomp scatter-gather walker
   - Remove request chaining
   - Add optional async request allocation

  Hashing:
   - Remove request chaining
   - Add optional async request allocation
   - Move partial block handling into API
   - Add ahash support to hmac
   - Fix shash documentation to disallow usage in hard IRQs

  Algorithms:
   - Remove unnecessary SIMD fallback code on x86 and arm/arm64
   - Drop avx10_256 xts(aes)/ctr(aes) on x86
   - Improve avx-512 optimisations for xts(aes)
   - Move chacha arch implementations into lib/crypto
   - Move poly1305 into lib/crypto and drop unused Crypto API algorithm
   - Disable powerpc/poly1305 as it has no SIMD fallback
   - Move sha256 arch implementations into lib/crypto
   - Convert deflate to acomp
   - Set block size correctly in cbcmac

  Drivers:
   - Do not use sg_dma_len before mapping in sun8i-ss
   - Fix warm-reboot failure by making shutdown do more work in qat
   - Add locking in zynqmp-sha
   - Remove cavium/zip
   - Add support for PCI device 0x17D8 to ccp
   - Add qat_6xxx support in qat
   - Add support for RK3576 in rockchip-rng
   - Add support for i.MX8QM in caam

  Others:
   - Fix irq_fpu_usable/kernel_fpu_begin inconsistency during CPU bring-up
   - Add new SEV/SNP platform shutdown API in ccp"

* tag 'v6.16-p1' of git://git.kernel.org/pub/scm/linux/kernel/git/herbert/crypto-2.6: (382 commits)
  x86/fpu: Fix irq_fpu_usable() to return false during CPU onlining
  crypto: qat - add missing header inclusion
  crypto: api - Redo lookup on EEXIST
  Revert "crypto: testmgr - Add hash export format testing"
  crypto: marvell/cesa - Do not chain submitted requests
  crypto: powerpc/poly1305 - add depends on BROKEN for now
  Revert "crypto: powerpc/poly1305 - Add SIMD fallback"
  crypto: ccp - Add missing tee info reg for teev2
  crypto: ccp - Add missing bootloader info reg for pspv5
  crypto: sun8i-ce - move fallback ahash_request to the end of the struct
  crypto: octeontx2 - Use dynamic allocated memory region for lmtst
  crypto: octeontx2 - Initialize cptlfs device info once
  crypto: xts - Only add ecb if it is not already there
  crypto: lrw - Only add ecb if it is not already there
  crypto: testmgr - Add hash export format testing
  crypto: testmgr - Use ahash for generic tfm
  crypto: hmac - Add ahash support
  crypto: testmgr - Ignore EEXIST on shash allocation
  crypto: algapi - Add driver template support to crypto_inst_setname
  crypto: shash - Set reqsize in shash_alg
  ...

1 files changed, 70 insertions, 0 deletions

diff --git a/arch/powerpc/lib/crypto/sha256.c b/arch/powerpc/lib/crypto/sha256.c
new file mode 100644
index 000000000000..6b0f079587eb
--- /dev/null
+++ b/arch/powerpc/lib/crypto/sha256.c
@@ -0,0 +1,70 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * SHA-256 Secure Hash Algorithm, SPE optimized
+ *
+ * Based on generic implementation. The assembler module takes care
+ * about the SPE registers so it can run from interrupt context.
+ *
+ * Copyright (c) 2015 Markus Stockhausen <stockhausen@collogia.de>
+ */
+
+#include <asm/switch_to.h>
+#include <crypto/internal/sha2.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/preempt.h>
+
+/*
+ * MAX_BYTES defines the number of bytes that are allowed to be processed
+ * between preempt_disable() and preempt_enable(). SHA256 takes ~2,000
+ * operations per 64 bytes. e500 cores can issue two arithmetic instructions
+ * per clock cycle using one 32/64 bit unit (SU1) and one 32 bit unit (SU2).
+ * Thus 1KB of input data will need an estimated maximum of 18,000 cycles.
+ * Headroom for cache misses included. Even with the low end model clocked
+ * at 667 MHz this equals to a critical time window of less than 27us.
+ *
+ */
+#define MAX_BYTES 1024
+
+extern void ppc_spe_sha256_transform(u32 *state, const u8 *src, u32 blocks);
+
+static void spe_begin(void)
+{
+	/* We just start SPE operations and will save SPE registers later. */
+	preempt_disable();
+	enable_kernel_spe();
+}
+
+static void spe_end(void)
+{
+	disable_kernel_spe();
+	/* reenable preemption */
+	preempt_enable();
+}
+
+void sha256_blocks_arch(u32 state[SHA256_STATE_WORDS],
+			const u8 *data, size_t nblocks)
+{
+	do {
+		/* cut input data into smaller blocks */
+		u32 unit = min_t(size_t, nblocks,
+				 MAX_BYTES / SHA256_BLOCK_SIZE);
+
+		spe_begin();
+		ppc_spe_sha256_transform(state, data, unit);
+		spe_end();
+
+		data += unit * SHA256_BLOCK_SIZE;
+		nblocks -= unit;
+	} while (nblocks);
+}
+EXPORT_SYMBOL_GPL(sha256_blocks_arch);
+
+bool sha256_is_arch_optimized(void)
+{
+	return true;
+}
+EXPORT_SYMBOL_GPL(sha256_is_arch_optimized);
+
+MODULE_LICENSE("GPL");
+MODULE_DESCRIPTION("SHA-256 Secure Hash Algorithm, SPE optimized");