Re: [RFC PATCH 32/42] docker: Add gentoo-mipsr5900el-cross image

[Philippe Mathieu-Daudé]

On 3/12/21 7:24 PM, Fredrik Noring wrote:
> On Friday, 12 March 2021, Maciej W. Rozycki wrote:
>> On Fri, 12 Mar 2021, Philippe Mathieu-Daudé wrote:
>>
>>>>>> Is there any way we can do this with a distro that isn't Gentoo
>>>>>> so that we can get a container build that is fast enough to be
>>>>>> useful for CI ?
>>>
>>> Using the Debian cross image I get:
>>>
>>> /home/phil/source/qemu/tests/docker/docker.py --engine auto cc --cc
>>> mips64el-linux-gnuabi64-gcc -i qemu/debian-mips64el-cross -s
>>> /home/phil/source/qemu --   -Wall -Werror -O0 -g -fno-strict-aliasing
>>> -mabi=n32 -march=r5900
>>> /home/phil/source/qemu/tests/tcg/mips/test-r5900-dmult.c -o
>>> test-r5900-dmult  -static
>>> cc1: error: unsupported combination: -march=r5900 -mhard-float
>>> -mdouble-float
>>>
>>> No clue what is setting '-mhard-float -mdouble-float' yet.
>>
>>  The R5900 has an FPU that only supports the single floating-point format.  
>> It's also not an IEEE 754 format.  The Linux kernel ABI does support the 
>> double and also the single floating-point format, both compliant with IEEE 
>> 754.
>>
>>  In the absence of a suitable FPU emulation code included with the kernel 
>> will handle the missing instructions (you can use the `nofpu' kernel 
>> parameter to force that in the presence of an FPU too).  Beware however 
>> that a recent change to the Linux kernel made FPU emulation code optional 
>> to suit some deeply embedded applications known never to use FPU machine 
>> instructions.
>>
>>  NB the presence of emulation is always required for MIPS ISA compliance 
>> if FPU machine instructions are ever to be used in a given application, 
>> because operations are allowed to trap regardless and rely on emulation.
>>
>>  I don't know what you are trying to achieve,
> 
> I believe Philippe is trying to compile
> 
> https://lists.gnu.org/archive/html/qemu-devel/2021-03/msg04565.html
> 
> testing this:
> 
>       The R5900 reports itself as MIPS III but does not implement DMULT.
>       Verify that DMULT is emulated properly in user mode by multiplying
>       two 64-bit numbers to produce a 128-bit number.
> 
> with this piece of code (notice the mips3 ISA directive for DMULT):
> 
> /*
>  * Test DMULT.
>  */
> 
> #include <stdio.h>
> #include <inttypes.h>
> #include <assert.h>
> 
> struct hi_lo { int64_t hi; uint64_t lo; };
> 
> static struct hi_lo dmult(int64_t rs, int64_t rt)
> {
>     int64_t hi;
>     uint64_t lo;
> 
>     /*
>      * The R5900 reports itself as MIPS III but does not implement DMULT.
>      * Verify that DMULT is emulated properly in user mode.
>      */
>     __asm__ __volatile__ (
>             "    .set  mips3\n"
>             "    dmult %2, %3\n"
>             "    mfhi  %0\n"
>             "    mflo  %1\n"
>             : "=r" (hi), "=r" (lo)
>             : "r" (rs), "r" (rt));
> 
>     return (struct hi_lo) { .hi = hi, .lo = lo };
> }
> 
> int main()
> {
>     /* Verify that multiplying two 64-bit numbers yields a 128-bit number. */
>     struct hi_lo r = dmult(2760727302517, 5665449960167);
> 
>     assert(r.hi == 847887);
>     assert(r.lo == 7893651516417804947);
> 
>     return 0;
> }
> 
>> but your two options to choose from are:
>>
>> 1. Build for the soft-float ABI (`-msoft-float') where any FP calculations 
>>    are compiled such as to be made by the CPU using integer arithmetic.
>>
>> 2. Build for a generic MIPS ISA, for the R5900/n32 that would be MIPS III
>>    (`-march=mips3'), and rely on the kernel FPU emulation.  Note that some 
>>    integer MIPS III operations are missing too from the R5900 and have to 
>>    be emulated by the kernel for MIPS/Linux n32 psABI compliance (an 
>>    implementation can be pinched from an old libgcc version that was still 
>>    under GNU GPLv2 or another algorithm reused, e.g. my `__div64_32' piece
>>    easily adapted).
> 
> So qemu/tests/tcg/mips/Makefile.target is patched with
> 
> # r5900 is only 64 bit little-endian
> ifneq ($(findstring 64el,$(TARGET_NAME)),)
> MIPS_TESTS += test-r5900-dmult
> test-r5900-dmult: CFLAGS += -mabi=n32 -march=r5900
> endif
> 
> I didn't have issues with the -mhard-float -mdouble-float flags at the time,
> and I didn't use mips64el-linux-gnuabi64-gcc for Debian.

Yes, this works with Gentoo. But building the Gentoo's r5900 toolchain
container is too expensive for our CI, so I was trying with Debian, but
with no luck so far.

I'll test the other image your suggested elsewhere and see if Daniel is
OK with it (I doubt because this is not an "official" image).

Thanks,

Phil.