Re: [PATCH 0/7] MIPS: math-emu: Branch delay slot emulation fixes
From: Aurelien Jarno <aurelien@aurel32.net>
Date: 2016-01-23 12:39:34
Hi, On 2016-01-22 05:20, Maciej W. Rozycki wrote:
Hi, This small patch series addresses the following issues with branch delay slot emulation in our floating-point emulator: - NOP emulation sometimes causes SIGILL (Aurelien's bug), - microMIPS emulation always goes astray, - microMIPS emulation of ADDIUPC always returns the wrong result. Also included are a bunch of code clean-ups and comment fixes. See individual patch descriptions for further details. I attempted to move clean-ups to the end, so that they do not interfere with backporting, except with 2/7 which, if reordered, would require 3/7 to become ill-formatted. I hope this is OK. Changes 5-7/7 do not require backporting. This series has been validated with a MIPS M5150 processor.
Thanks for working on that. I have tested this series, and I confirm this fixes the issue and works well, though I haven't tested them on a system with microMIPS support. So for all the patches: Reviewed-by: Aurelien Jarno <aurelien@aurel32.net> And for patch 1: Tested-by: Aurelien Jarno <aurelien@aurel32.net> Aurelien -- Aurelien Jarno GPG: 4096R/1DDD8C9B aurelien@aurel32.net http://www.aurel32.net