lightning
[Top][All Lists]
Advanced

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

Re: Gnu-Lightning test suite failures on armv7h


From: Zachary Cook
Subject: Re: Gnu-Lightning test suite failures on armv7h
Date: Mon, 10 Feb 2020 15:37:49 -0500

This test was run on an ODROID XU4 running Arch Linux Arm

When I compile with --enable-assertions all tests fail due to:
lt-lightning: jit_disasm.c:143: jit_init_debug: Assertion `disasm_print' failed.

cpuinfo:
processor : 0
model name : ARMv7 Processor rev 3 (v7l)
BogoMIPS : 12.00
Features : half thumb fastmult vfp edsp neon vfpv3 tls vfpv4 idiva idivt vfpd32 lpae
CPU implementer : 0x41
CPU architecture: 7
CPU variant : 0x0
CPU part : 0xc07
CPU revision : 3

On Mon, Feb 10, 2020 at 1:40 PM Paulo César Pereira de Andrade <address@hidden> wrote:
Em sex., 24 de jan. de 2020 às 17:01, Zachary Cook
<address@hidden> escreveu:

  Hi Zachary,

> See https://gist.github.com/ZachCook/767a115a87e2ec3ae2e7ff1969ca4147 for test suite log, same results on both 2.1.3 and git master

  What environment are you using?

  I am using an almost 6 years old qemu image as arm test environment, and
all tests pass for me.

  Since it is failing only when generating arm instructions, not for thumb
instructions, it should be something specific to arm instructions. Either
some kind of bug, or maybe it is some kind of environment that only knows
about thumb instructions... Quick googling it should be the case, based on
https://en.wikipedia.org/wiki/ARM_Cortex-M:
"""
Only Thumb-1 and Thumb-2 instruction sets are supported in Cortex-M
architectures; the legacy 32-bit ARM instruction set isn't supported.
"""
  What do you see as 'cat /proc/cpuinfo' output? It should be patched to
just skip tests for arm instructions on this environment.

Thanks,
Paulo

reply via email to

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