public inbox for gcc-cvs@sourceware.org
help / color / mirror / Atom feed
* [gcc r12-6765] arm: suppress aes erratum when forwarding from aes
@ 2022-01-20 11:29 Richard Earnshaw
0 siblings, 0 replies; only message in thread
From: Richard Earnshaw @ 2022-01-20 11:29 UTC (permalink / raw)
To: gcc-cvs
https://gcc.gnu.org/g:2078550a005f3fde4c331ad4b8452c963c4cdb9d
commit r12-6765-g2078550a005f3fde4c331ad4b8452c963c4cdb9d
Author: Richard Earnshaw <rearnsha@arm.com>
Date: Mon Nov 1 13:23:26 2021 +0000
arm: suppress aes erratum when forwarding from aes
AES operations are commonly chained and since the result of one AES
operation is never a 32-bit value, they do not need an additional
mitigation instruction for the forwarded result. We handle this
common case by adding additional patterns that allow for this.
gcc/ChangeLog:
* config/arm/crypto.md (crypto_<CRYPTO_AESMC:crypto_pattern>_protected):
New pattern.
(aarch32_crypto_aese_fused_protected): Likewise.
(aarch32_crypto_aesd_fused_protected): Likewise.
Diff:
---
gcc/config/arm/crypto.md | 50 ++++++++++++++++++++++++++++++++++++++++++++++++
1 file changed, 50 insertions(+)
diff --git a/gcc/config/arm/crypto.md b/gcc/config/arm/crypto.md
index fbee1829ce8..df857352382 100644
--- a/gcc/config/arm/crypto.md
+++ b/gcc/config/arm/crypto.md
@@ -75,6 +75,20 @@
[(set_attr "type" "neon_move_q")]
)
+;; An AESMC operation can feed directly into a subsequent AES
+;; operation without needing mitigation.
+(define_insn "*crypto_<CRYPTO_AESMC:crypto_pattern>_protected"
+ [(set (match_operand:<crypto_mode> 0 "register_operand" "=w")
+ (unspec:<crypto_mode>
+ [(unspec:<crypto_mode>
+ [(match_operand:<crypto_mode> 1 "register_operand" "w")]
+ CRYPTO_AESMC)]
+ UNSPEC_AES_PROTECT))]
+ "TARGET_CRYPTO && fix_aes_erratum_1742098"
+ "<crypto_pattern>.<crypto_size_sfx>\\t%q0, %q1"
+ [(set_attr "type" "<crypto_type>")]
+)
+
;; When AESE/AESMC fusion is enabled we really want to keep the two together
;; and enforce the register dependency without scheduling or register
;; allocation messing up the order or introducing moves inbetween.
@@ -95,6 +109,25 @@
(set_attr "length" "8")]
)
+;; And similarly when mitigation is enabled, but not needed in this
+;; case.
+(define_insn "*aarch32_crypto_aese_fused_protected"
+ [(set (match_operand:V16QI 0 "register_operand" "=w")
+ (unspec:V16QI
+ [(unspec:V16QI
+ [(unspec:V16QI [(xor:V16QI
+ (match_operand:V16QI 1 "register_operand" "%0")
+ (match_operand:V16QI 2 "register_operand" "w"))]
+ UNSPEC_AESE)]
+ UNSPEC_AESMC)]
+ UNSPEC_AES_PROTECT))]
+ "TARGET_CRYPTO && fix_aes_erratum_1742098
+ && arm_fusion_enabled_p (tune_params::FUSE_AES_AESMC)"
+ "aese.8\\t%q0, %q2\;aesmc.8\\t%q0, %q0"
+ [(set_attr "type" "crypto_aese")
+ (set_attr "length" "8")]
+)
+
;; When AESD/AESIMC fusion is enabled we really want to keep the two together
;; and enforce the register dependency without scheduling or register
;; allocation messing up the order or introducing moves inbetween.
@@ -115,6 +148,23 @@
(set_attr "length" "8")]
)
+(define_insn "*aarch32_crypto_aesd_fused_protected"
+ [(set (match_operand:V16QI 0 "register_operand" "=w")
+ (unspec:V16QI
+ [(unspec:V16QI
+ [(unspec:V16QI [(xor:V16QI
+ (match_operand:V16QI 1 "register_operand" "%0")
+ (match_operand:V16QI 2 "register_operand" "w"))]
+ UNSPEC_AESD)]
+ UNSPEC_AESIMC)]
+ UNSPEC_AES_PROTECT))]
+ "TARGET_CRYPTO && fix_aes_erratum_1742098
+ && arm_fusion_enabled_p (tune_params::FUSE_AES_AESMC)"
+ "aesd.8\\t%q0, %q2\;aesimc.8\\t%q0, %q0"
+ [(set_attr "type" "crypto_aese")
+ (set_attr "length" "8")]
+)
+
(define_insn "crypto_<CRYPTO_BINARY:crypto_pattern>"
[(set (match_operand:<crypto_mode> 0 "register_operand" "=w")
(unspec:<crypto_mode>
^ permalink raw reply [flat|nested] only message in thread
only message in thread, other threads:[~2022-01-20 11:29 UTC | newest]
Thread overview: (only message) (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2022-01-20 11:29 [gcc r12-6765] arm: suppress aes erratum when forwarding from aes Richard Earnshaw
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).