Fix for Random failures in System.Numerics.Tests.modpowTest.ModPowAxiom test #74112
Conversation
ghost
added
area-System.Numerics
community-contribution
|
Tagging subscribers to this area: @dotnet/area-system-numerics |
danmoseley
changed the title
|
Updated the title -- issue numbers in titles don't work well when looking at lists of PR's 😸 |
|
.NET 7 port needed? |
|
@danmoseley , I I think yes, .NET 7 port is needed because without the fix |
dakersnar
left a comment
dakersnar
left a comment
There was a problem hiding this comment.
Looks good to me, nice catch. Thanks for the fix!
| power >>= 1; | ||
| } | ||
|
|
||
| return result.Slice(0, resultLength); |
There was a problem hiding this comment.
Was this slice basically unnecessary in all cases, and causing problems in this edge case?
There was a problem hiding this comment.
Yes, correct. It doesn't exist for overload of PowCore with exponent of arbitrary length.
There was a problem hiding this comment.
Removing this slice introduced a regression for the case of:
value: 174475926811173371108797680334274448482027679060635889276326576767485928691470272416521788770673419883076040463212449631887959313694345475337324466914914839368969988215395980439434099840700753576299320684517028939674493640310781015739837159625716992654178036326745036841619965661654305787656989089177185300627
exponent: 152266387011071825363527751452194927195391511201743678326712300944812006670853968954038119879530327815990523885596782666640476470053822766345969220504425389995206265094671158964311234812226260664473986637197447367726400093404731405598815953979171704175885394570388678952091206779039984338579147114424357147650
modulus: 15548277136038299103
There was a problem hiding this comment.
In this scenario, the length of result is greater than the length of bits and so the result.CopyTo(bits) operation fails as the destination is too short.
I expect the actual issue here is that resultLength isn't being correctly tracked somewhere.
There was a problem hiding this comment.
Its not that resultLength is incorrect, its that bits will have been mutated and may contain temporary data
src/libraries/System.Runtime.Numerics/src/System/Numerics/BigIntegerCalculator.PowMod.cs
Show resolved
Hide resolved
|
@danmoseley Yes, I'd like to get this in for .NET 7, as this is a regression from .NET 6. |
|
/backport to release/7.0-rc1 |
|
Started backporting to release/7.0-rc1: https://github.com/dotnet/runtime/actions/runs/2885517952 |
ghost
locked as resolved and limited conversation to collaborators
4 participants
Fixes #70330