diff options
author | René van Dorst <opensource@vdorst.com> | 2018-09-21 16:53:01 +0200 |
---|---|---|
committer | Jason A. Donenfeld <Jason@zx2c4.com> | 2018-09-22 02:56:07 +0200 |
commit | 1e4a1352ce4d4e6bc883e3e520e62d1d00c9f909 (patch) | |
tree | bfd56268a66c1459150587f9cdf6da697000df29 | |
parent | da44b52039c58f73ae25b45598105ca7ef3fe275 (diff) |
chacha20-mips32r2: reduce jumptable entry size and stack usage
Signed-off-by: René van Dorst <opensource@vdorst.com>
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
-rw-r--r-- | src/crypto/zinc/chacha20/chacha20-mips.S | 87 |
1 files changed, 48 insertions, 39 deletions
diff --git a/src/crypto/zinc/chacha20/chacha20-mips.S b/src/crypto/zinc/chacha20/chacha20-mips.S index 7e2b5e8..8796da3 100644 --- a/src/crypto/zinc/chacha20/chacha20-mips.S +++ b/src/crypto/zinc/chacha20/chacha20-mips.S @@ -49,7 +49,7 @@ * They are used to handling the last bytes which are not multiple of 4. */ #define SAVED_X X15 -#define SAVED_CA $ra +#define SAVED_CA $fp #define PTR_LAST_ROUND $v1 #define IS_UNALIGNED $fp @@ -166,25 +166,23 @@ CONCAT3(.Lchacha20_mips_xor_aligned_, PLUS_ONE(x), _b: ;) \ #define JMPTBL_ALIGNED(x) \ .Lchacha20_mips_jmptbl_aligned_ ## x: ; \ .set noreorder; \ + b .Lchacha20_mips_xor_aligned_ ## x ## _b; \ .if (x == 12); \ - move SAVED_CA, NONCE_0; \ + addu SAVED_X, X ## x, NONCE_0; \ .else; \ - lw SAVED_CA, (x*4)(STATE); \ + addu SAVED_X, X ## x, SAVED_CA; \ .endif; \ - b .Lchacha20_mips_xor_aligned_ ## x ## _b; \ - move SAVED_X, X ## x; \ .set reorder #define JMPTBL_UNALIGNED(x) \ .Lchacha20_mips_jmptbl_unaligned_ ## x: ; \ .set noreorder; \ + b .Lchacha20_mips_xor_unaligned_ ## x ## _b; \ .if (x == 12); \ - move SAVED_CA, NONCE_0; \ + addu SAVED_X, X ## x, NONCE_0; \ .else; \ - lw SAVED_CA, (x*4)(STATE);\ + addu SAVED_X, X ## x, SAVED_CA; \ .endif; \ - b .Lchacha20_mips_xor_unaligned_ ## x ## _b; \ - move SAVED_X, X ## x; \ .set reorder #define AXR(A, B, C, D, K, L, M, N, V, W, Y, Z, S) \ @@ -219,17 +217,16 @@ chacha20_mips: ins PTR_LAST_ROUND, $zero, 0, 6 addu PTR_LAST_ROUND, OUT - /* Save s0-s7, ra, fp */ - sw $ra, 0($sp) - sw $fp, 4($sp) - sw $s0, 8($sp) - sw $s1, 12($sp) - sw $s2, 16($sp) - sw $s3, 20($sp) - sw $s4, 24($sp) - sw $s5, 28($sp) - sw $s6, 32($sp) - sw $s7, 36($sp) + /* Save s0-s7, fp */ + sw $s0, 0($sp) + sw $s1, 4($sp) + sw $s2, 8($sp) + sw $s3, 12($sp) + sw $s4, 16($sp) + sw $s5, 20($sp) + sw $s6, 24($sp) + sw $s7, 28($sp) + sw $fp, 32($sp) lw NONCE_0, 48(STATE) @@ -293,31 +290,45 @@ chacha20_mips: /* Last round? No, do a full block. */ bne OUT, PTR_LAST_ROUND, .Lchacha20_mips_xor_aligned_16_b - /* Calculate lower half jump table addr and offset */ - ins T0, $at, 2, 6 + /* Calculate lower half jump table offset */ + ins T0, $at, 1, 6 /* Full block? Jump */ beqz BYTES, .Lchacha20_mips_xor_aligned_16_b - subu T0, $at + /* Add STATE with offset */ + addu T1, STATE, $at + + /* Add lower half jump table addr */ addiu T0, %lo(.Lchacha20_mips_jmptbl_aligned_0) + + /* Read value from STATE */ + lw SAVED_CA, 0(T1) + jr T0 .Loop_chacha20_unaligned: /* Load upper half of jump table addr */ lui T0, %hi(.Lchacha20_mips_jmptbl_unaligned_0) - /* Last round? no jump */ + /* Last round? No, do a full block. */ bne OUT, PTR_LAST_ROUND, .Lchacha20_mips_xor_unaligned_16_b - /* Calculate lower half jump table addr and offset */ - ins T0, $at, 2, 6 + /* Calculate lower half jump table offset */ + ins T0, $at, 1, 6 /* Full block? Jump */ beqz BYTES, .Lchacha20_mips_xor_unaligned_16_b - subu T0, $at + /* Add STATE with offset */ + addu T1, STATE, $at + + /* Add lower half jump table addr */ addiu T0, %lo(.Lchacha20_mips_jmptbl_unaligned_0) + + /* Read value from STATE */ + lw SAVED_CA, 0(T1) + jr T0 /* Aligned code path @@ -336,16 +347,15 @@ chacha20_mips: sw NONCE_0, 48(STATE) /* Restore used registers */ - lw $ra, 0($sp) - lw $fp, 4($sp) - lw $s0, 8($sp) - lw $s1, 12($sp) - lw $s2, 16($sp) - lw $s3, 20($sp) - lw $s4, 24($sp) - lw $s5, 28($sp) - lw $s6, 32($sp) - lw $s7, 36($sp) + lw $s0, 0($sp) + lw $s1, 4($sp) + lw $s2, 8($sp) + lw $s3, 12($sp) + lw $s4, 16($sp) + lw $s5, 20($sp) + lw $s6, 24($sp) + lw $s7, 28($sp) + lw $fp, 32($sp) .Lchacha20_mips_end: addiu $sp, STACK_SIZE jr $ra @@ -368,9 +378,8 @@ chacha20_mips: .set reorder .Lchacha20_mips_xor_bytes: - addu OUT, $at addu IN, $at - addu SAVED_X, SAVED_CA + addu OUT, $at /* First byte */ lbu T1, 0(IN) andi $at, BYTES, 2 |