Minki/linux
1
0
Fork 1
mirror of https://github.com/torvalds/linux.git synced 2024-11-16 00:52:01 +00:00
Code Issues Packages Projects Releases Wiki Activity
18fe6bf505
linux/arch/x86/crypto/sha1_ni_asm.S
Linus Torvalds 94a855111e - Add the call depth tracking mitigation for Retbleed which has 
been long in the making. It is a lighterweight software-only fix for
 Skylake-based cores where enabling IBRS is a big hammer and causes a
 significant performance impact.
 
 What it basically does is, it aligns all kernel functions to 16 bytes
 boundary and adds a 16-byte padding before the function, objtool
 collects all functions' locations and when the mitigation gets applied,
 it patches a call accounting thunk which is used to track the call depth
 of the stack at any time.
 
 When that call depth reaches a magical, microarchitecture-specific value
 for the Return Stack Buffer, the code stuffs that RSB and avoids its
 underflow which could otherwise lead to the Intel variant of Retbleed.
 
 This software-only solution brings a lot of the lost performance back,
 as benchmarks suggest:
 
   https://lore.kernel.org/all/20220915111039.092790446@infradead.org/
 
 That page above also contains a lot more detailed explanation of the
 whole mechanism
 
 - Implement a new control flow integrity scheme called FineIBT which is
 based on the software kCFI implementation and uses hardware IBT support
 where present to annotate and track indirect branches using a hash to
 validate them
 
 - Other misc fixes and cleanups
 -----BEGIN PGP SIGNATURE-----
 
 iQIzBAABCgAdFiEEzv7L6UO9uDPlPSfHEsHwGGHeVUoFAmOZp5EACgkQEsHwGGHe
 VUrZFxAAvi/+8L0IYSK4mKJvixGbTFjxN/Swo2JVOfs34LqGUT6JaBc+VUMwZxdb
 VMTFIZ3ttkKEodjhxGI7oGev6V8UfhI37SmO2lYKXpQVjXXnMlv/M+Vw3teE38CN
 gopi+xtGnT1IeWQ3tc/Tv18pleJ0mh5HKWiW+9KoqgXj0wgF9x4eRYDz1TDCDA/A
 iaBzs56j8m/FSykZHnrWZ/MvjKNPdGlfJASUCPeTM2dcrXQGJ93+X2hJctzDte0y
 Nuiw6Y0htfFBE7xoJn+sqm5Okr+McoUM18/CCprbgSKYk18iMYm3ZtAi6FUQZS1A
 ua4wQCf49loGp15PO61AS5d3OBf5D3q/WihQRbCaJvTVgPp9sWYnWwtcVUuhMllh
 ZQtBU9REcVJ/22bH09Q9CjBW0VpKpXHveqQdqRDViLJ6v/iI6EFGmD24SW/VxyRd
 73k9MBGrL/dOf1SbEzdsnvcSB3LGzp0Om8o/KzJWOomrVKjBCJy16bwTEsCZEJmP
 i406m92GPXeaN1GhTko7vmF0GnkEdJs1GVCZPluCAxxbhHukyxHnrjlQjI4vC80n
 Ylc0B3Kvitw7LGJsPqu+/jfNHADC/zhx1qz/30wb5cFmFbN1aRdp3pm8JYUkn+l/
 zri2Y6+O89gvE/9/xUhMohzHsWUO7xITiBavewKeTP9GSWybWUs=
 =cRy1
 -----END PGP SIGNATURE-----

Merge tag 'x86_core_for_v6.2' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip

Pull x86 core updates from Borislav Petkov:

 - Add the call depth tracking mitigation for Retbleed which has been
   long in the making. It is a lighterweight software-only fix for
   Skylake-based cores where enabling IBRS is a big hammer and causes a
   significant performance impact.

   What it basically does is, it aligns all kernel functions to 16 bytes
   boundary and adds a 16-byte padding before the function, objtool
   collects all functions' locations and when the mitigation gets
   applied, it patches a call accounting thunk which is used to track
   the call depth of the stack at any time.

   When that call depth reaches a magical, microarchitecture-specific
   value for the Return Stack Buffer, the code stuffs that RSB and
   avoids its underflow which could otherwise lead to the Intel variant
   of Retbleed.

   This software-only solution brings a lot of the lost performance
   back, as benchmarks suggest:

       https://lore.kernel.org/all/20220915111039.092790446@infradead.org/

   That page above also contains a lot more detailed explanation of the
   whole mechanism

 - Implement a new control flow integrity scheme called FineIBT which is
   based on the software kCFI implementation and uses hardware IBT
   support where present to annotate and track indirect branches using a
   hash to validate them

 - Other misc fixes and cleanups

* tag 'x86_core_for_v6.2' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (80 commits)
  x86/paravirt: Use common macro for creating simple asm paravirt functions
  x86/paravirt: Remove clobber bitmask from .parainstructions
  x86/debug: Include percpu.h in debugreg.h to get DECLARE_PER_CPU() et al
  x86/cpufeatures: Move X86_FEATURE_CALL_DEPTH from bit 18 to bit 19 of word 11, to leave space for WIP X86_FEATURE_SGX_EDECCSSA bit
  x86/Kconfig: Enable kernel IBT by default
  x86,pm: Force out-of-line memcpy()
  objtool: Fix weak hole vs prefix symbol
  objtool: Optimize elf_dirty_reloc_sym()
  x86/cfi: Add boot time hash randomization
  x86/cfi: Boot time selection of CFI scheme
  x86/ibt: Implement FineIBT
  objtool: Add --cfi to generate the .cfi_sites section
  x86: Add prefix symbols for function padding
  objtool: Add option to generate prefix symbols
  objtool: Avoid O(bloody terrible) behaviour -- an ode to libelf
  objtool: Slice up elf_create_section_symbol()
  kallsyms: Revert "Take callthunks into account"
  x86: Unconfuse CONFIG_ and X86_FEATURE_ namespaces
  x86/retpoline: Fix crash printing warning
  x86/paravirt: Fix a !PARAVIRT build warning
  ...
2022-12-14 15:03:00 -08:00

305 lines
7.7 KiB
ArmAsm
Raw Blame History

/*
* Intel SHA Extensions optimized implementation of a SHA-1 update function
*
* This file is provided under a dual BSD/GPLv2 license. When using or
* redistributing this file, you may do so under either license.
*
* GPL LICENSE SUMMARY
*
* Copyright(c) 2015 Intel Corporation.
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of version 2 of the GNU General Public License as
* published by the Free Software Foundation.
*
* This program is distributed in the hope that it will be useful, but
* WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
* General Public License for more details.
*
* Contact Information:
* Sean Gulley <sean.m.gulley@intel.com>
* Tim Chen <tim.c.chen@linux.intel.com>
*
* BSD LICENSE
*
* Copyright(c) 2015 Intel Corporation.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
*
* * Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* * Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in
* the documentation and/or other materials provided with the
* distribution.
* * Neither the name of Intel Corporation nor the names of its
* contributors may be used to endorse or promote products derived
* from this software without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
* "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
* LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
* A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
* OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
* SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
* LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
* DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
* THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
* (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
* OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*
*/
#include <linux/linkage.h>
#include <linux/cfi_types.h>
#define DIGEST_PTR %rdi /* 1st arg */
#define DATA_PTR %rsi /* 2nd arg */
#define NUM_BLKS %rdx /* 3rd arg */
/* gcc conversion */
#define FRAME_SIZE 32 /* space for 2x16 bytes */
#define ABCD %xmm0
#define E0 %xmm1 /* Need two E's b/c they ping pong */
#define E1 %xmm2
#define MSG0 %xmm3
#define MSG1 %xmm4
#define MSG2 %xmm5
#define MSG3 %xmm6
#define SHUF_MASK %xmm7
/*
* Intel SHA Extensions optimized implementation of a SHA-1 update function
*
* The function takes a pointer to the current hash values, a pointer to the
* input data, and a number of 64 byte blocks to process. Once all blocks have
* been processed, the digest pointer is updated with the resulting hash value.
* The function only processes complete blocks, there is no functionality to
* store partial blocks. All message padding and hash value initialization must
* be done outside the update function.
*
* The indented lines in the loop are instructions related to rounds processing.
* The non-indented lines are instructions related to the message schedule.
*
* void sha1_ni_transform(uint32_t *digest, const void *data,
uint32_t numBlocks)
* digest : pointer to digest
* data: pointer to input data
* numBlocks: Number of blocks to process
*/
.text
SYM_TYPED_FUNC_START(sha1_ni_transform)
push %rbp
mov %rsp, %rbp
sub $FRAME_SIZE, %rsp
and $~0xF, %rsp
shl $6, NUM_BLKS /* convert to bytes */
jz .Ldone_hash
add DATA_PTR, NUM_BLKS /* pointer to end of data */
/* load initial hash values */
pinsrd $3, 1*16(DIGEST_PTR), E0
movdqu 0*16(DIGEST_PTR), ABCD
pand UPPER_WORD_MASK(%rip), E0
pshufd $0x1B, ABCD, ABCD
movdqa PSHUFFLE_BYTE_FLIP_MASK(%rip), SHUF_MASK
.Lloop0:
/* Save hash values for addition after rounds */
movdqa E0, (0*16)(%rsp)
movdqa ABCD, (1*16)(%rsp)
/* Rounds 0-3 */
movdqu 0*16(DATA_PTR), MSG0
pshufb SHUF_MASK, MSG0
paddd MSG0, E0
movdqa ABCD, E1
sha1rnds4 $0, E0, ABCD
/* Rounds 4-7 */
movdqu 1*16(DATA_PTR), MSG1
pshufb SHUF_MASK, MSG1
sha1nexte MSG1, E1
movdqa ABCD, E0
sha1rnds4 $0, E1, ABCD
sha1msg1 MSG1, MSG0
/* Rounds 8-11 */
movdqu 2*16(DATA_PTR), MSG2
pshufb SHUF_MASK, MSG2
sha1nexte MSG2, E0
movdqa ABCD, E1
sha1rnds4 $0, E0, ABCD
sha1msg1 MSG2, MSG1
pxor MSG2, MSG0
/* Rounds 12-15 */
movdqu 3*16(DATA_PTR), MSG3
pshufb SHUF_MASK, MSG3
sha1nexte MSG3, E1
movdqa ABCD, E0
sha1msg2 MSG3, MSG0
sha1rnds4 $0, E1, ABCD
sha1msg1 MSG3, MSG2
pxor MSG3, MSG1
/* Rounds 16-19 */
sha1nexte MSG0, E0
movdqa ABCD, E1
sha1msg2 MSG0, MSG1
sha1rnds4 $0, E0, ABCD
sha1msg1 MSG0, MSG3
pxor MSG0, MSG2
/* Rounds 20-23 */
sha1nexte MSG1, E1
movdqa ABCD, E0
sha1msg2 MSG1, MSG2
sha1rnds4 $1, E1, ABCD
sha1msg1 MSG1, MSG0
pxor MSG1, MSG3
/* Rounds 24-27 */
sha1nexte MSG2, E0
movdqa ABCD, E1
sha1msg2 MSG2, MSG3
sha1rnds4 $1, E0, ABCD
sha1msg1 MSG2, MSG1
pxor MSG2, MSG0
/* Rounds 28-31 */
sha1nexte MSG3, E1
movdqa ABCD, E0
sha1msg2 MSG3, MSG0
sha1rnds4 $1, E1, ABCD
sha1msg1 MSG3, MSG2
pxor MSG3, MSG1
/* Rounds 32-35 */
sha1nexte MSG0, E0
movdqa ABCD, E1
sha1msg2 MSG0, MSG1
sha1rnds4 $1, E0, ABCD
sha1msg1 MSG0, MSG3
pxor MSG0, MSG2
/* Rounds 36-39 */
sha1nexte MSG1, E1
movdqa ABCD, E0
sha1msg2 MSG1, MSG2
sha1rnds4 $1, E1, ABCD
sha1msg1 MSG1, MSG0
pxor MSG1, MSG3
/* Rounds 40-43 */
sha1nexte MSG2, E0
movdqa ABCD, E1
sha1msg2 MSG2, MSG3
sha1rnds4 $2, E0, ABCD
sha1msg1 MSG2, MSG1
pxor MSG2, MSG0
/* Rounds 44-47 */
sha1nexte MSG3, E1
movdqa ABCD, E0
sha1msg2 MSG3, MSG0
sha1rnds4 $2, E1, ABCD
sha1msg1 MSG3, MSG2
pxor MSG3, MSG1
/* Rounds 48-51 */
sha1nexte MSG0, E0
movdqa ABCD, E1
sha1msg2 MSG0, MSG1
sha1rnds4 $2, E0, ABCD
sha1msg1 MSG0, MSG3
pxor MSG0, MSG2
/* Rounds 52-55 */
sha1nexte MSG1, E1
movdqa ABCD, E0
sha1msg2 MSG1, MSG2
sha1rnds4 $2, E1, ABCD
sha1msg1 MSG1, MSG0
pxor MSG1, MSG3
/* Rounds 56-59 */
sha1nexte MSG2, E0
movdqa ABCD, E1
sha1msg2 MSG2, MSG3
sha1rnds4 $2, E0, ABCD
sha1msg1 MSG2, MSG1
pxor MSG2, MSG0
/* Rounds 60-63 */
sha1nexte MSG3, E1
movdqa ABCD, E0
sha1msg2 MSG3, MSG0
sha1rnds4 $3, E1, ABCD
sha1msg1 MSG3, MSG2
pxor MSG3, MSG1
/* Rounds 64-67 */
sha1nexte MSG0, E0
movdqa ABCD, E1
sha1msg2 MSG0, MSG1
sha1rnds4 $3, E0, ABCD
sha1msg1 MSG0, MSG3
pxor MSG0, MSG2
/* Rounds 68-71 */
sha1nexte MSG1, E1
movdqa ABCD, E0
sha1msg2 MSG1, MSG2
sha1rnds4 $3, E1, ABCD
pxor MSG1, MSG3
/* Rounds 72-75 */
sha1nexte MSG2, E0
movdqa ABCD, E1
sha1msg2 MSG2, MSG3
sha1rnds4 $3, E0, ABCD
/* Rounds 76-79 */
sha1nexte MSG3, E1
movdqa ABCD, E0
sha1rnds4 $3, E1, ABCD
/* Add current hash values with previously saved */
sha1nexte (0*16)(%rsp), E0
paddd (1*16)(%rsp), ABCD
/* Increment data pointer and loop if more to process */
add $64, DATA_PTR
cmp NUM_BLKS, DATA_PTR
jne .Lloop0
/* Write hash values back in the correct order */
pshufd $0x1B, ABCD, ABCD
movdqu ABCD, 0*16(DIGEST_PTR)
pextrd $3, E0, 1*16(DIGEST_PTR)
.Ldone_hash:
mov %rbp, %rsp
pop %rbp
RET
SYM_FUNC_END(sha1_ni_transform)
.section .rodata.cst16.PSHUFFLE_BYTE_FLIP_MASK, "aM", @progbits, 16
.align 16
PSHUFFLE_BYTE_FLIP_MASK:
.octa 0x000102030405060708090a0b0c0d0e0f
.section .rodata.cst16.UPPER_WORD_MASK, "aM", @progbits, 16
.align 16
UPPER_WORD_MASK:
.octa 0xFFFFFFFF000000000000000000000000
Reference in New Issue View Git Blame Copy Permalink