Skip to content

Comments

Fix a carry overflow bug in bn_sqr_comba4/8 for mips 32-bit targets#17258

Closed
bernd-edlinger wants to merge 1 commit intoopenssl:masterfrom
bernd-edlinger:fix_overflow_bug_in_bn_srq_comba
Closed

Fix a carry overflow bug in bn_sqr_comba4/8 for mips 32-bit targets#17258
bernd-edlinger wants to merge 1 commit intoopenssl:masterfrom
bernd-edlinger:fix_overflow_bug_in_bn_srq_comba

Conversation

@bernd-edlinger
Copy link
Member

@bernd-edlinger bernd-edlinger commented Dec 12, 2021

bn_sqr_comba8 does for instance compute a wrong result for the value:
a=0x4aaac919 62056c84 fba7334e 1a6be678 022181ba fd3aa878 899b2346 ee210f45

The correct result is:
r=0x15c72e32 605a3061 d11b1012 3c187483 6df96999 bd0c22ba d3e7d437 4724a82f 912c5e61 6a187efe 8f7c47fc f6945fe5 75be8e3d 97ed17d4 7950b4653 cb32899

but the actual result was:
r=0x15c72e32 605a3061 d11b1012 3c187483 6df96999 bd0c22ba d3e7d437 4724a82f 912c5e61 6a187efe 8f7c47fc f6945fe5 75be8e3c 97ed17d4 7950b4653 cb32899

so the forth word of the result was 0x75be8e3c but should have been
0x75be8e3d instead.

Likewise bn_sqr_comba4 has an identical bug for the same value as well:
a=0x022181ba fd3aa878 899b2346 ee210f45

correct result:
r=0x00048a69 9fe82f8b 62bd2ed1 88781335 75be8e3d 97ed17d4 7950b465 3cb3289

wrong result:
r=0x00048a69 9fe82f8b 62bd2ed1 88781335 75be8e3c 97ed17d4 7950b465 3cb3289

Fortunately the bn_mul_comba4/8 code paths are not affected.

Although the same assembler source is used on mips64 targets it does
in fact handle the carry propagation correctly, as far as I can tell.

As a work around we disable the BN_SQR_COMBA feature on the affected targets,
until the assembler source can be fixed.

Checklist
  • documentation is added or updated
  • tests are added or updated

@bernd-edlinger bernd-edlinger added branch: master Applies to master branch branch: 1.0.2 Applies to OpenSSL_1_0_2-stable branch (EOL) branch: 1.1.1 Applies to OpenSSL_1_1_1-stable branch (EOL) branch: 3.0 Applies to openssl-3.0 branch labels Dec 12, 2021
@github-actions github-actions bot added the severity: fips change The pull request changes FIPS provider sources label Dec 12, 2021
@bernd-edlinger bernd-edlinger force-pushed the fix_overflow_bug_in_bn_srq_comba branch 3 times, most recently from 35b2a63 to efd0db6 Compare December 12, 2021 15:50
@bernd-edlinger
Copy link
Member Author

Okay, after some testing, I'm confident I have understood the problem in the assembler now,
and therefore replaced this PR to fix the mips squaring code instead of disabling it.

bn_sqr_comba8 does for instance compute a wrong result for the value:
a=0x4aaac919 62056c84 fba7334e 1a6be678 022181ba fd3aa878 899b2346 ee210f45

The correct result is:
r=0x15c72e32 605a3061 d11b1012 3c187483 6df96999 bd0c22ba d3e7d437 4724a82f
    912c5e61 6a187efe 8f7c47fc f6945fe5 75be8e3d 97ed17d4 7950b465 3cb3289

but the actual result was:
r=0x15c72e32 605a3061 d11b1012 3c187483 6df96999 bd0c22ba d3e7d437 4724a82f
    912c5e61 6a187efe 8f7c47fc f6945fe5 75be8e3c 97ed17d4 7950b465 3cb3289

so the forth word of the result was 0x75be8e3c but should have been
0x75be8e3d instead.

Likewise bn_sqr_comba4 has an identical bug for the same value as well:
a=0x022181ba fd3aa878 899b2346 ee210f45

correct result:
r=0x00048a69 9fe82f8b 62bd2ed1 88781335 75be8e3d 97ed17d4 7950b465 3cb3289

wrong result:
r=0x00048a69 9fe82f8b 62bd2ed1 88781335 75be8e3c 97ed17d4 7950b465 3cb3289

Fortunately the bn_mul_comba4/8 code paths are not affected.

Also the mips64 target does in fact not handle the carry propagation
correctly.

Example:
a=0x4aaac91900000000 62056c8400000000 fba7334e00000000 1a6be67800000000
    022181ba00000000 fd3aa87800000000 899b234635dad283 ee210f4500000001

correct result:
r=0x15c72e32272c4471 392debf018c679c8 b85496496bf8254c d0204f36611e2be1
    0cdb3db8f3c081d8 c94ba0e1bacc5061 191b83d47ff929f6 5be0aebfc13ae68d
    3eea7a7fdf2f5758 42f7ec656cab3cb5 6a28095be34756f2 64f24687bf37de06
    2822309cd1d292f9 6fa698c972372f09 771e97d3a868cda0 dc421e8a00000001

wrong result:
r=0x15c72e32272c4471 392debf018c679c8 b85496496bf8254c d0204f36611e2be1
    0cdb3db8f3c081d8 c94ba0e1bacc5061 191b83d47ff929f6 5be0aebfc13ae68d
    3eea7a7fdf2f5758 42f7ec656cab3cb5 6a28095be34756f2 64f24687bf37de06
    2822309cd1d292f8 6fa698c972372f09 771e97d3a868cda0 dc421e8a00000001
@bernd-edlinger bernd-edlinger force-pushed the fix_overflow_bug_in_bn_srq_comba branch from efd0db6 to 43972e1 Compare December 12, 2021 16:04
@paulidale paulidale added the approval: done This pull request has the required number of approvals label Dec 12, 2021
@bernd-edlinger
Copy link
Member Author

24h rule is fulfilled, and this is actually urgent,
therefore setting to ready-to-merge

@bernd-edlinger bernd-edlinger added approval: ready to merge The 24 hour grace period has passed, ready to merge and removed approval: done This pull request has the required number of approvals labels Dec 14, 2021
openssl-machine pushed a commit that referenced this pull request Dec 14, 2021
bn_sqr_comba8 does for instance compute a wrong result for the value:
a=0x4aaac919 62056c84 fba7334e 1a6be678 022181ba fd3aa878 899b2346 ee210f45

The correct result is:
r=0x15c72e32 605a3061 d11b1012 3c187483 6df96999 bd0c22ba d3e7d437 4724a82f
    912c5e61 6a187efe 8f7c47fc f6945fe5 75be8e3d 97ed17d4 7950b465 3cb3289

but the actual result was:
r=0x15c72e32 605a3061 d11b1012 3c187483 6df96999 bd0c22ba d3e7d437 4724a82f
    912c5e61 6a187efe 8f7c47fc f6945fe5 75be8e3c 97ed17d4 7950b465 3cb3289

so the forth word of the result was 0x75be8e3c but should have been
0x75be8e3d instead.

Likewise bn_sqr_comba4 has an identical bug for the same value as well:
a=0x022181ba fd3aa878 899b2346 ee210f45

correct result:
r=0x00048a69 9fe82f8b 62bd2ed1 88781335 75be8e3d 97ed17d4 7950b465 3cb3289

wrong result:
r=0x00048a69 9fe82f8b 62bd2ed1 88781335 75be8e3c 97ed17d4 7950b465 3cb3289

Fortunately the bn_mul_comba4/8 code paths are not affected.

Also the mips64 target does in fact not handle the carry propagation
correctly.

Example:
a=0x4aaac91900000000 62056c8400000000 fba7334e00000000 1a6be67800000000
    022181ba00000000 fd3aa87800000000 899b234635dad283 ee210f4500000001

correct result:
r=0x15c72e32272c4471 392debf018c679c8 b85496496bf8254c d0204f36611e2be1
    0cdb3db8f3c081d8 c94ba0e1bacc5061 191b83d47ff929f6 5be0aebfc13ae68d
    3eea7a7fdf2f5758 42f7ec656cab3cb5 6a28095be34756f2 64f24687bf37de06
    2822309cd1d292f9 6fa698c972372f09 771e97d3a868cda0 dc421e8a00000001

wrong result:
r=0x15c72e32272c4471 392debf018c679c8 b85496496bf8254c d0204f36611e2be1
    0cdb3db8f3c081d8 c94ba0e1bacc5061 191b83d47ff929f6 5be0aebfc13ae68d
    3eea7a7fdf2f5758 42f7ec656cab3cb5 6a28095be34756f2 64f24687bf37de06
    2822309cd1d292f8 6fa698c972372f09 771e97d3a868cda0 dc421e8a00000001

Reviewed-by: Paul Dale <[email protected]>
(Merged from #17258)
openssl-machine pushed a commit that referenced this pull request Dec 14, 2021
bn_sqr_comba8 does for instance compute a wrong result for the value:
a=0x4aaac919 62056c84 fba7334e 1a6be678 022181ba fd3aa878 899b2346 ee210f45

The correct result is:
r=0x15c72e32 605a3061 d11b1012 3c187483 6df96999 bd0c22ba d3e7d437 4724a82f
    912c5e61 6a187efe 8f7c47fc f6945fe5 75be8e3d 97ed17d4 7950b465 3cb3289

but the actual result was:
r=0x15c72e32 605a3061 d11b1012 3c187483 6df96999 bd0c22ba d3e7d437 4724a82f
    912c5e61 6a187efe 8f7c47fc f6945fe5 75be8e3c 97ed17d4 7950b465 3cb3289

so the forth word of the result was 0x75be8e3c but should have been
0x75be8e3d instead.

Likewise bn_sqr_comba4 has an identical bug for the same value as well:
a=0x022181ba fd3aa878 899b2346 ee210f45

correct result:
r=0x00048a69 9fe82f8b 62bd2ed1 88781335 75be8e3d 97ed17d4 7950b465 3cb3289

wrong result:
r=0x00048a69 9fe82f8b 62bd2ed1 88781335 75be8e3c 97ed17d4 7950b465 3cb3289

Fortunately the bn_mul_comba4/8 code paths are not affected.

Also the mips64 target does in fact not handle the carry propagation
correctly.

Example:
a=0x4aaac91900000000 62056c8400000000 fba7334e00000000 1a6be67800000000
    022181ba00000000 fd3aa87800000000 899b234635dad283 ee210f4500000001

correct result:
r=0x15c72e32272c4471 392debf018c679c8 b85496496bf8254c d0204f36611e2be1
    0cdb3db8f3c081d8 c94ba0e1bacc5061 191b83d47ff929f6 5be0aebfc13ae68d
    3eea7a7fdf2f5758 42f7ec656cab3cb5 6a28095be34756f2 64f24687bf37de06
    2822309cd1d292f9 6fa698c972372f09 771e97d3a868cda0 dc421e8a00000001

wrong result:
r=0x15c72e32272c4471 392debf018c679c8 b85496496bf8254c d0204f36611e2be1
    0cdb3db8f3c081d8 c94ba0e1bacc5061 191b83d47ff929f6 5be0aebfc13ae68d
    3eea7a7fdf2f5758 42f7ec656cab3cb5 6a28095be34756f2 64f24687bf37de06
    2822309cd1d292f8 6fa698c972372f09 771e97d3a868cda0 dc421e8a00000001

Reviewed-by: Paul Dale <[email protected]>
(Merged from #17258)

(cherry picked from commit 336923c)
openssl-machine pushed a commit that referenced this pull request Dec 14, 2021
bn_sqr_comba8 does for instance compute a wrong result for the value:
a=0x4aaac919 62056c84 fba7334e 1a6be678 022181ba fd3aa878 899b2346 ee210f45

The correct result is:
r=0x15c72e32 605a3061 d11b1012 3c187483 6df96999 bd0c22ba d3e7d437 4724a82f
    912c5e61 6a187efe 8f7c47fc f6945fe5 75be8e3d 97ed17d4 7950b465 3cb3289

but the actual result was:
r=0x15c72e32 605a3061 d11b1012 3c187483 6df96999 bd0c22ba d3e7d437 4724a82f
    912c5e61 6a187efe 8f7c47fc f6945fe5 75be8e3c 97ed17d4 7950b465 3cb3289

so the forth word of the result was 0x75be8e3c but should have been
0x75be8e3d instead.

Likewise bn_sqr_comba4 has an identical bug for the same value as well:
a=0x022181ba fd3aa878 899b2346 ee210f45

correct result:
r=0x00048a69 9fe82f8b 62bd2ed1 88781335 75be8e3d 97ed17d4 7950b465 3cb3289

wrong result:
r=0x00048a69 9fe82f8b 62bd2ed1 88781335 75be8e3c 97ed17d4 7950b465 3cb3289

Fortunately the bn_mul_comba4/8 code paths are not affected.

Also the mips64 target does in fact not handle the carry propagation
correctly.

Example:
a=0x4aaac91900000000 62056c8400000000 fba7334e00000000 1a6be67800000000
    022181ba00000000 fd3aa87800000000 899b234635dad283 ee210f4500000001

correct result:
r=0x15c72e32272c4471 392debf018c679c8 b85496496bf8254c d0204f36611e2be1
    0cdb3db8f3c081d8 c94ba0e1bacc5061 191b83d47ff929f6 5be0aebfc13ae68d
    3eea7a7fdf2f5758 42f7ec656cab3cb5 6a28095be34756f2 64f24687bf37de06
    2822309cd1d292f9 6fa698c972372f09 771e97d3a868cda0 dc421e8a00000001

wrong result:
r=0x15c72e32272c4471 392debf018c679c8 b85496496bf8254c d0204f36611e2be1
    0cdb3db8f3c081d8 c94ba0e1bacc5061 191b83d47ff929f6 5be0aebfc13ae68d
    3eea7a7fdf2f5758 42f7ec656cab3cb5 6a28095be34756f2 64f24687bf37de06
    2822309cd1d292f8 6fa698c972372f09 771e97d3a868cda0 dc421e8a00000001

Reviewed-by: Paul Dale <[email protected]>
(Merged from #17258)

(cherry picked from commit 336923c)
@bernd-edlinger
Copy link
Member Author

Merged to all branches. Thanks!

bernd-edlinger added a commit to bernd-edlinger/openssl that referenced this pull request Dec 14, 2021
bn_sqr_comba8 does for instance compute a wrong result for the value:
a=0x4aaac919 62056c84 fba7334e 1a6be678 022181ba fd3aa878 899b2346 ee210f45

The correct result is:
r=0x15c72e32 605a3061 d11b1012 3c187483 6df96999 bd0c22ba d3e7d437 4724a82f
    912c5e61 6a187efe 8f7c47fc f6945fe5 75be8e3d 97ed17d4 7950b465 3cb3289

but the actual result was:
r=0x15c72e32 605a3061 d11b1012 3c187483 6df96999 bd0c22ba d3e7d437 4724a82f
    912c5e61 6a187efe 8f7c47fc f6945fe5 75be8e3c 97ed17d4 7950b465 3cb3289

so the forth word of the result was 0x75be8e3c but should have been
0x75be8e3d instead.

Likewise bn_sqr_comba4 has an identical bug for the same value as well:
a=0x022181ba fd3aa878 899b2346 ee210f45

correct result:
r=0x00048a69 9fe82f8b 62bd2ed1 88781335 75be8e3d 97ed17d4 7950b465 3cb3289

wrong result:
r=0x00048a69 9fe82f8b 62bd2ed1 88781335 75be8e3c 97ed17d4 7950b465 3cb3289

Fortunately the bn_mul_comba4/8 code paths are not affected.

Also the mips64 target does in fact not handle the carry propagation
correctly.

Example:
a=0x4aaac91900000000 62056c8400000000 fba7334e00000000 1a6be67800000000
    022181ba00000000 fd3aa87800000000 899b234635dad283 ee210f4500000001

correct result:
r=0x15c72e32272c4471 392debf018c679c8 b85496496bf8254c d0204f36611e2be1
    0cdb3db8f3c081d8 c94ba0e1bacc5061 191b83d47ff929f6 5be0aebfc13ae68d
    3eea7a7fdf2f5758 42f7ec656cab3cb5 6a28095be34756f2 64f24687bf37de06
    2822309cd1d292f9 6fa698c972372f09 771e97d3a868cda0 dc421e8a00000001

wrong result:
r=0x15c72e32272c4471 392debf018c679c8 b85496496bf8254c d0204f36611e2be1
    0cdb3db8f3c081d8 c94ba0e1bacc5061 191b83d47ff929f6 5be0aebfc13ae68d
    3eea7a7fdf2f5758 42f7ec656cab3cb5 6a28095be34756f2 64f24687bf37de06
    2822309cd1d292f8 6fa698c972372f09 771e97d3a868cda0 dc421e8a00000001

Reviewed-by: Paul Dale <[email protected]>
(Merged from openssl#17258)

(cherry picked from commit 336923c)
bernd-edlinger added a commit to bernd-edlinger/openssl that referenced this pull request Dec 14, 2021
bn_sqr_comba8 does for instance compute a wrong result for the value:
a=0x4aaac919 62056c84 fba7334e 1a6be678 022181ba fd3aa878 899b2346 ee210f45

The correct result is:
r=0x15c72e32 605a3061 d11b1012 3c187483 6df96999 bd0c22ba d3e7d437 4724a82f
    912c5e61 6a187efe 8f7c47fc f6945fe5 75be8e3d 97ed17d4 7950b465 3cb3289

but the actual result was:
r=0x15c72e32 605a3061 d11b1012 3c187483 6df96999 bd0c22ba d3e7d437 4724a82f
    912c5e61 6a187efe 8f7c47fc f6945fe5 75be8e3c 97ed17d4 7950b465 3cb3289

so the forth word of the result was 0x75be8e3c but should have been
0x75be8e3d instead.

Likewise bn_sqr_comba4 has an identical bug for the same value as well:
a=0x022181ba fd3aa878 899b2346 ee210f45

correct result:
r=0x00048a69 9fe82f8b 62bd2ed1 88781335 75be8e3d 97ed17d4 7950b465 3cb3289

wrong result:
r=0x00048a69 9fe82f8b 62bd2ed1 88781335 75be8e3c 97ed17d4 7950b465 3cb3289

Fortunately the bn_mul_comba4/8 code paths are not affected.

Also the mips64 target does in fact not handle the carry propagation
correctly.

Example:
a=0x4aaac91900000000 62056c8400000000 fba7334e00000000 1a6be67800000000
    022181ba00000000 fd3aa87800000000 899b234635dad283 ee210f4500000001

correct result:
r=0x15c72e32272c4471 392debf018c679c8 b85496496bf8254c d0204f36611e2be1
    0cdb3db8f3c081d8 c94ba0e1bacc5061 191b83d47ff929f6 5be0aebfc13ae68d
    3eea7a7fdf2f5758 42f7ec656cab3cb5 6a28095be34756f2 64f24687bf37de06
    2822309cd1d292f9 6fa698c972372f09 771e97d3a868cda0 dc421e8a00000001

wrong result:
r=0x15c72e32272c4471 392debf018c679c8 b85496496bf8254c d0204f36611e2be1
    0cdb3db8f3c081d8 c94ba0e1bacc5061 191b83d47ff929f6 5be0aebfc13ae68d
    3eea7a7fdf2f5758 42f7ec656cab3cb5 6a28095be34756f2 64f24687bf37de06
    2822309cd1d292f8 6fa698c972372f09 771e97d3a868cda0 dc421e8a00000001

Reviewed-by: Paul Dale <[email protected]>
(Merged from openssl#17258)

(cherry picked from commit 336923c)
dongbeiouba pushed a commit to auvkone/BabaSSL that referenced this pull request Jan 29, 2022
bn_sqr_comba8 does for instance compute a wrong result for the value:
a=0x4aaac919 62056c84 fba7334e 1a6be678 022181ba fd3aa878 899b2346 ee210f45

The correct result is:
r=0x15c72e32 605a3061 d11b1012 3c187483 6df96999 bd0c22ba d3e7d437 4724a82f
    912c5e61 6a187efe 8f7c47fc f6945fe5 75be8e3d 97ed17d4 7950b465 3cb32899

but the actual result was:
r=0x15c72e32 605a3061 d11b1012 3c187483 6df96999 bd0c22ba d3e7d437 4724a82f
    912c5e61 6a187efe 8f7c47fc f6945fe5 75be8e3c 97ed17d4 7950b465 3cb32899

so the forth word of the result was 0x75be8e3c but should have been
0x75be8e3d instead.

Likewise bn_sqr_comba4 has an identical bug for the same value as well:
a=0x022181ba fd3aa878 899b2346 ee210f45

correct result:
r=0x00048a69 9fe82f8b 62bd2ed1 88781335 75be8e3d 97ed17d4 7950b465 3cb32899

wrong result:
r=0x00048a69 9fe82f8b 62bd2ed1 88781335 75be8e3c 97ed17d4 7950b465 3cb32899

Fortunately the bn_mul_comba4/8 code paths are not affected.

Also the mips64 target does in fact not handle the carry propagation
correctly.

Example:
a=0x4aaac91900000000 62056c8400000000 fba7334e00000000 1a6be67800000000
    022181ba00000000 fd3aa87800000000 899b234635dad283 ee210f4500000001

correct result:
r=0x15c72e32272c4471 392debf018c679c8 b85496496bf8254c d0204f36611e2be1
    0cdb3db8f3c081d8 c94ba0e1bacc5061 191b83d47ff929f6 5be0aebfc13ae68d
    3eea7a7fdf2f5758 42f7ec656cab3cb5 6a28095be34756f2 64f24687bf37de06
    2822309cd1d292f9 6fa698c972372f09 771e97d3a868cda0 dc421e8a00000001

wrong result:
r=0x15c72e32272c4471 392debf018c679c8 b85496496bf8254c d0204f36611e2be1
    0cdb3db8f3c081d8 c94ba0e1bacc5061 191b83d47ff929f6 5be0aebfc13ae68d
    3eea7a7fdf2f5758 42f7ec656cab3cb5 6a28095be34756f2 64f24687bf37de06
    2822309cd1d292f8 6fa698c972372f09 771e97d3a868cda0 dc421e8a00000001

Reviewed-by: Paul Dale <[email protected]>
(Merged from openssl/openssl#17258)

(cherry picked from commit 336923c0c8d705cb8af5216b29a205662db0d590)
dongbeiouba pushed a commit to Tongsuo-Project/Tongsuo that referenced this pull request Feb 14, 2022
bn_sqr_comba8 does for instance compute a wrong result for the value:
a=0x4aaac919 62056c84 fba7334e 1a6be678 022181ba fd3aa878 899b2346 ee210f45

The correct result is:
r=0x15c72e32 605a3061 d11b1012 3c187483 6df96999 bd0c22ba d3e7d437 4724a82f
    912c5e61 6a187efe 8f7c47fc f6945fe5 75be8e3d 97ed17d4 7950b465 3cb32899

but the actual result was:
r=0x15c72e32 605a3061 d11b1012 3c187483 6df96999 bd0c22ba d3e7d437 4724a82f
    912c5e61 6a187efe 8f7c47fc f6945fe5 75be8e3c 97ed17d4 7950b465 3cb32899

so the forth word of the result was 0x75be8e3c but should have been
0x75be8e3d instead.

Likewise bn_sqr_comba4 has an identical bug for the same value as well:
a=0x022181ba fd3aa878 899b2346 ee210f45

correct result:
r=0x00048a69 9fe82f8b 62bd2ed1 88781335 75be8e3d 97ed17d4 7950b465 3cb32899

wrong result:
r=0x00048a69 9fe82f8b 62bd2ed1 88781335 75be8e3c 97ed17d4 7950b465 3cb32899

Fortunately the bn_mul_comba4/8 code paths are not affected.

Also the mips64 target does in fact not handle the carry propagation
correctly.

Example:
a=0x4aaac91900000000 62056c8400000000 fba7334e00000000 1a6be67800000000
    022181ba00000000 fd3aa87800000000 899b234635dad283 ee210f4500000001

correct result:
r=0x15c72e32272c4471 392debf018c679c8 b85496496bf8254c d0204f36611e2be1
    0cdb3db8f3c081d8 c94ba0e1bacc5061 191b83d47ff929f6 5be0aebfc13ae68d
    3eea7a7fdf2f5758 42f7ec656cab3cb5 6a28095be34756f2 64f24687bf37de06
    2822309cd1d292f9 6fa698c972372f09 771e97d3a868cda0 dc421e8a00000001

wrong result:
r=0x15c72e32272c4471 392debf018c679c8 b85496496bf8254c d0204f36611e2be1
    0cdb3db8f3c081d8 c94ba0e1bacc5061 191b83d47ff929f6 5be0aebfc13ae68d
    3eea7a7fdf2f5758 42f7ec656cab3cb5 6a28095be34756f2 64f24687bf37de06
    2822309cd1d292f8 6fa698c972372f09 771e97d3a868cda0 dc421e8a00000001

Reviewed-by: Paul Dale <[email protected]>
(Merged from openssl/openssl#17258)

(cherry picked from commit 336923c0c8d705cb8af5216b29a205662db0d590)
bernd-edlinger added a commit to bernd-edlinger/openssl that referenced this pull request Mar 16, 2022
bn_sqr_comba8 does for instance compute a wrong result for the value:
a=0x4aaac919 62056c84 fba7334e 1a6be678 022181ba fd3aa878 899b2346 ee210f45

The correct result is:
r=0x15c72e32 605a3061 d11b1012 3c187483 6df96999 bd0c22ba d3e7d437 4724a82f
    912c5e61 6a187efe 8f7c47fc f6945fe5 75be8e3d 97ed17d4 7950b465 3cb3289

but the actual result was:
r=0x15c72e32 605a3061 d11b1012 3c187483 6df96999 bd0c22ba d3e7d437 4724a82f
    912c5e61 6a187efe 8f7c47fc f6945fe5 75be8e3c 97ed17d4 7950b465 3cb3289

so the forth word of the result was 0x75be8e3c but should have been
0x75be8e3d instead.

Likewise bn_sqr_comba4 has an identical bug for the same value as well:
a=0x022181ba fd3aa878 899b2346 ee210f45

correct result:
r=0x00048a69 9fe82f8b 62bd2ed1 88781335 75be8e3d 97ed17d4 7950b465 3cb3289

wrong result:
r=0x00048a69 9fe82f8b 62bd2ed1 88781335 75be8e3c 97ed17d4 7950b465 3cb3289

Fortunately the bn_mul_comba4/8 code paths are not affected.

Also the mips64 target does in fact not handle the carry propagation
correctly.

Example:
a=0x4aaac91900000000 62056c8400000000 fba7334e00000000 1a6be67800000000
    022181ba00000000 fd3aa87800000000 899b234635dad283 ee210f4500000001

correct result:
r=0x15c72e32272c4471 392debf018c679c8 b85496496bf8254c d0204f36611e2be1
    0cdb3db8f3c081d8 c94ba0e1bacc5061 191b83d47ff929f6 5be0aebfc13ae68d
    3eea7a7fdf2f5758 42f7ec656cab3cb5 6a28095be34756f2 64f24687bf37de06
    2822309cd1d292f9 6fa698c972372f09 771e97d3a868cda0 dc421e8a00000001

wrong result:
r=0x15c72e32272c4471 392debf018c679c8 b85496496bf8254c d0204f36611e2be1
    0cdb3db8f3c081d8 c94ba0e1bacc5061 191b83d47ff929f6 5be0aebfc13ae68d
    3eea7a7fdf2f5758 42f7ec656cab3cb5 6a28095be34756f2 64f24687bf37de06
    2822309cd1d292f8 6fa698c972372f09 771e97d3a868cda0 dc421e8a00000001

Reviewed-by: Paul Dale <[email protected]>
(Merged from openssl#17258)

(cherry picked from commit 336923c)
raspbian-autopush pushed a commit to raspbian-packages/openssl that referenced this pull request Mar 17, 2022
bn_sqr_comba8 does for instance compute a wrong result for the value:
a=0x4aaac919 62056c84 fba7334e 1a6be678 022181ba fd3aa878 899b2346 ee210f45

The correct result is:
r=0x15c72e32 605a3061 d11b1012 3c187483 6df96999 bd0c22ba d3e7d437 4724a82f
    912c5e61 6a187efe 8f7c47fc f6945fe5 75be8e3d 97ed17d4 7950b465 3cb32899

but the actual result was:
r=0x15c72e32 605a3061 d11b1012 3c187483 6df96999 bd0c22ba d3e7d437 4724a82f
    912c5e61 6a187efe 8f7c47fc f6945fe5 75be8e3c 97ed17d4 7950b465 3cb32899

so the forth word of the result was 0x75be8e3c but should have been
0x75be8e3d instead.

Likewise bn_sqr_comba4 has an identical bug for the same value as well:
a=0x022181ba fd3aa878 899b2346 ee210f45

correct result:
r=0x00048a69 9fe82f8b 62bd2ed1 88781335 75be8e3d 97ed17d4 7950b465 3cb32899

wrong result:
r=0x00048a69 9fe82f8b 62bd2ed1 88781335 75be8e3c 97ed17d4 7950b465 3cb32899

Fortunately the bn_mul_comba4/8 code paths are not affected.

Also the mips64 target does in fact not handle the carry propagation
correctly.

Example:
a=0x4aaac91900000000 62056c8400000000 fba7334e00000000 1a6be67800000000
    022181ba00000000 fd3aa87800000000 899b234635dad283 ee210f4500000001

correct result:
r=0x15c72e32272c4471 392debf018c679c8 b85496496bf8254c d0204f36611e2be1
    0cdb3db8f3c081d8 c94ba0e1bacc5061 191b83d47ff929f6 5be0aebfc13ae68d
    3eea7a7fdf2f5758 42f7ec656cab3cb5 6a28095be34756f2 64f24687bf37de06
    2822309cd1d292f9 6fa698c972372f09 771e97d3a868cda0 dc421e8a00000001

wrong result:
r=0x15c72e32272c4471 392debf018c679c8 b85496496bf8254c d0204f36611e2be1
    0cdb3db8f3c081d8 c94ba0e1bacc5061 191b83d47ff929f6 5be0aebfc13ae68d
    3eea7a7fdf2f5758 42f7ec656cab3cb5 6a28095be34756f2 64f24687bf37de06
    2822309cd1d292f8 6fa698c972372f09 771e97d3a868cda0 dc421e8a00000001

Reviewed-by: Paul Dale <[email protected]>
(Merged from openssl/openssl#17258)

(cherry picked from commit 336923c0c8d705cb8af5216b29a205662db0d590)

Gbp-Pq: Name Fix-a-carry-overflow-bug-in-bn_sqr_comba4-8-for-mips-32-b.patch
raspbian-autopush pushed a commit to raspbian-packages/openssl that referenced this pull request Mar 17, 2022
bn_sqr_comba8 does for instance compute a wrong result for the value:
a=0x4aaac919 62056c84 fba7334e 1a6be678 022181ba fd3aa878 899b2346 ee210f45

The correct result is:
r=0x15c72e32 605a3061 d11b1012 3c187483 6df96999 bd0c22ba d3e7d437 4724a82f
    912c5e61 6a187efe 8f7c47fc f6945fe5 75be8e3d 97ed17d4 7950b465 3cb32899

but the actual result was:
r=0x15c72e32 605a3061 d11b1012 3c187483 6df96999 bd0c22ba d3e7d437 4724a82f
    912c5e61 6a187efe 8f7c47fc f6945fe5 75be8e3c 97ed17d4 7950b465 3cb32899

so the forth word of the result was 0x75be8e3c but should have been
0x75be8e3d instead.

Likewise bn_sqr_comba4 has an identical bug for the same value as well:
a=0x022181ba fd3aa878 899b2346 ee210f45

correct result:
r=0x00048a69 9fe82f8b 62bd2ed1 88781335 75be8e3d 97ed17d4 7950b465 3cb32899

wrong result:
r=0x00048a69 9fe82f8b 62bd2ed1 88781335 75be8e3c 97ed17d4 7950b465 3cb32899

Fortunately the bn_mul_comba4/8 code paths are not affected.

Also the mips64 target does in fact not handle the carry propagation
correctly.

Example:
a=0x4aaac91900000000 62056c8400000000 fba7334e00000000 1a6be67800000000
    022181ba00000000 fd3aa87800000000 899b234635dad283 ee210f4500000001

correct result:
r=0x15c72e32272c4471 392debf018c679c8 b85496496bf8254c d0204f36611e2be1
    0cdb3db8f3c081d8 c94ba0e1bacc5061 191b83d47ff929f6 5be0aebfc13ae68d
    3eea7a7fdf2f5758 42f7ec656cab3cb5 6a28095be34756f2 64f24687bf37de06
    2822309cd1d292f9 6fa698c972372f09 771e97d3a868cda0 dc421e8a00000001

wrong result:
r=0x15c72e32272c4471 392debf018c679c8 b85496496bf8254c d0204f36611e2be1
    0cdb3db8f3c081d8 c94ba0e1bacc5061 191b83d47ff929f6 5be0aebfc13ae68d
    3eea7a7fdf2f5758 42f7ec656cab3cb5 6a28095be34756f2 64f24687bf37de06
    2822309cd1d292f8 6fa698c972372f09 771e97d3a868cda0 dc421e8a00000001

Reviewed-by: Paul Dale <[email protected]>
(Merged from openssl/openssl#17258)

(cherry picked from commit 336923c0c8d705cb8af5216b29a205662db0d590)

Gbp-Pq: Name Fix-a-carry-overflow-bug-in-bn_sqr_comba4-8-for-mips-32-b.patch