lightning
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[PATCH 0/6] Misc patches to pass the testsuite and remove warnings


From: Kim Kuparinen
Subject: [PATCH 0/6] Misc patches to pass the testsuite and remove warnings
Date: Sat, 20 Nov 2021 13:13:21 +0200

Resend of patches I previously sent without signatures, along with two misc
additions. Sorry about the extra spam.

I've tested these patches with Debian 11, kernel Linux 5.10.70-1 on qemu on
arches

ppc64le, cpu POWER9
mips64el, cpu 5KEc
arm, cpu cortex-a15
aarch64, cpu cortex-a57
riscv64

and on real hardware with amd64 (AMD Ryzen 5 5600x) and arm on my Raspberry Pi
4B (cortex-a72).

make check seems to pass on all systems tested, except qemu-system-ppc64le and
my Pi, although I'm pretty sure there's a bug in qemu-ppc64le. It has to do 
with the
fcti* instructions, which according to the specs I found should return either 
max/min
value of the target register when dealing with values larger/smaller than that, 
but qemu
only returns a `nanvalue`. When I ran qemu with a patch I wrote, ppc64le also
passes all tests.

My Pi passes when generating thumb code, but fails when generating arm code. No
clue what's going on, I get a segfault on the first generated instruction of 
every
test. I'll probably write a more specific bug report in the near future, but I'd
like to do some more testing on my own before then.

Kim Kuparinen (6):
  riscv: Fix movi/relatives loading incorrect value
  arm: Force RES0 bits in TST/A1 to be 0
  mips: Use DEXT when in 64-bit mode instead of EXT
  ppc: Add endianness check to _truncr_d_i
  misc: Add bfd_arch to arm/aarch64/mips
  Include math.h in check/lightning.c

 check/lightning.c   |  1 +
 lib/jit_arm-cpu.c   |  2 +-
 lib/jit_disasm.c    | 13 +++++++++++++
 lib/jit_mips-cpu.c  |  6 +++++-
 lib/jit_ppc-fpu.c   |  4 ++++
 lib/jit_riscv-cpu.c | 12 ++++++++++++
 6 files changed, 36 insertions(+), 2 deletions(-)

-- 
2.33.0




reply via email to

[Prev in Thread] Current Thread [Next in Thread]