[crypto][NFC] Fix compiler warnings about variable types in libcrypt
Needs ReviewPublic
Actions

Authored by gfunni234_gmail.com on Aug 22 2021, 6:51 PM.

Details

Reviewers

imp
jrtc27
dim

Summary

Instead of casting from unsigned and signed and vice versa, we should be consistent with signedness.

NFC this if possible please!

Diff Detail

Repository

rG FreeBSD src repository

Lint

Lint Skipped

Unit

Tests Skipped

Event Timeline

gfunni234_gmail.com requested review of this revision.Aug 22 2021, 6:51 PM

gfunni234_gmail.com created this revision.

gfunni234_gmail.com added reviewers: jrtc27, dim.Aug 24 2021, 3:53 PM

gfunni234_gmail.com updated this revision to Diff 94118.Aug 24 2021, 5:46 PM

gfunni234_gmail.com retitled this revision from [crypto][NFC]Fix compiler warnings about variable types in libcrypt to [crypto][NFC] Fix compiler warnings about variable types in libcrypt.

gfunni234_gmail.com edited the summary of this revision. (Show Details)

Most of these changes have nothing to do with what your summary says

lib/libcrypt/crypt-md5.c
46	So now we have a magic constant rather than being automatically calculated by strlen (which, yes, compilers know about and will constant evaluate)?
lib/libcrypt/crypt-sha256.c
186	This regresses code style