https://bugs.kde.org/show_bug.cgi?id=385334
Carl Love changed:
What|Removed |Added
Attachment #108173|0 |1
is obsolete||
https://bugs.kde.org/show_bug.cgi?id=385334
--- Comment #6 from Julian Seward ---
Random observation, w.r.t new patch:
Instead of this
binop( Iop_64HLtoV128,
mkU64( 0x1F1F1F1F1F1F1F1F ),
mkU64( 0x1
https://bugs.kde.org/show_bug.cgi?id=385334
Carl Love changed:
What|Removed |Added
Attachment #108147|0 |1
is obsolete||
https://bugs.kde.org/show_bug.cgi?id=385334
--- Comment #4 from Carl Love ---
The issue is with the vpermr and xxpermr instructions that were reimplemented
based on the vperm instruction. In those implementations, the mask needed to
be 0x1F instead of 0xF and the mask assignment uses perm_val or
https://bugs.kde.org/show_bug.cgi?id=385334
--- Comment #3 from Carl Love ---
I will go back and look at this again. The issue was actually seen in the
re-working of the vpermr, xxperm, xxpermr. The code from the vperm was used to
implement the other instructions with the needed tweaks to do th
https://bugs.kde.org/show_bug.cgi?id=385334
--- Comment #2 from Julian Seward ---
I looked at the IR generation for vperm
(case 0x2B: { // vperm (Permute, AV p218), etc)
and it looks correct to me. Which test case is failing and
for which target? (32-be, 64-be, 64-le) ?
The code already does us
https://bugs.kde.org/show_bug.cgi?id=385334
--- Comment #1 from Carl Love ---
Created attachment 108147
--> https://bugs.kde.org/attachment.cgi?id=108147&action=edit
vperm instuction, fix index mask
Add patch to fix issue
--
You are receiving this mail because:
You are watching all bug chang
https://bugs.kde.org/show_bug.cgi?id=385334
Carl Love changed:
What|Removed |Added
Ever confirmed|0 |1
Status|UNCONFIRMED |C
