Based-on: <20220607204557.658541-1-richard.henderson@linaro.org>
("[PATCH v4 00/53] semihosting cleanup")

Changes for v4:
  * Split out of v2.
  * Convert host errno to gdb errno, which for m68k is guest errno.


r~


Richard Henderson (3):
  target/m68k: Eliminate m68k_semi_is_fseek
  target/m68k: Make semihosting system only
  target/m68k: Use semihosting/syscalls.h

 linux-user/m68k/cpu_loop.c |   5 -
 target/m68k/m68k-semi.c    | 375 +++++++++----------------------------
 target/m68k/meson.build    |   6 +-
 3 files changed, 92 insertions(+), 294 deletions(-)

-- 
2.34.1

Le 08/06/2022 à 00:26, Richard Henderson a écrit :
> Based-on: <20220607204557.658541-1-richard.henderson@linaro.org>
> ("[PATCH v4 00/53] semihosting cleanup")
> 
> Changes for v4:
>    * Split out of v2.
>    * Convert host errno to gdb errno, which for m68k is guest errno.
> 

How do you test semihosting on m68k?

Thanks,
Laurent

On 6/8/22 02:36, Laurent Vivier wrote:
> Le 08/06/2022 à 00:26, Richard Henderson a écrit :
>> Based-on: <20220607204557.658541-1-richard.henderson@linaro.org>
>> ("[PATCH v4 00/53] semihosting cleanup")
>>
>> Changes for v4:
>>    * Split out of v2.
>>    * Convert host errno to gdb errno, which for m68k is guest errno.
>>
> 
> How do you test semihosting on m68k?

I have only compiled this.  I had been working on adding something to tests/tcg/m68k/, 
which is where those halt patches came from, but haven't finished.


r~

Le 08/06/2022 à 16:53, Richard Henderson a écrit :
> On 6/8/22 02:36, Laurent Vivier wrote:
>> Le 08/06/2022 à 00:26, Richard Henderson a écrit :
>>> Based-on: <20220607204557.658541-1-richard.henderson@linaro.org>
>>> ("[PATCH v4 00/53] semihosting cleanup")
>>>
>>> Changes for v4:
>>>    * Split out of v2.
>>>    * Convert host errno to gdb errno, which for m68k is guest errno.
>>>
>>
>> How do you test semihosting on m68k?
> 
> I have only compiled this.  I had been working on adding something to tests/tcg/m68k/, which is 
> where those halt patches came from, but haven't finished.
> 

But what is the use case of semihosting on m68k?
Is there a kernel implementation available for that?

Thanks,
Laurent

On 6/8/22 11:08, Laurent Vivier wrote:
> Le 08/06/2022 à 16:53, Richard Henderson a écrit :
>> On 6/8/22 02:36, Laurent Vivier wrote:
>>> Le 08/06/2022 à 00:26, Richard Henderson a écrit :
>>>> Based-on: <20220607204557.658541-1-richard.henderson@linaro.org>
>>>> ("[PATCH v4 00/53] semihosting cleanup")
>>>>
>>>> Changes for v4:
>>>>    * Split out of v2.
>>>>    * Convert host errno to gdb errno, which for m68k is guest errno.
>>>>
>>>
>>> How do you test semihosting on m68k?
>>
>> I have only compiled this.  I had been working on adding something to tests/tcg/m68k/, 
>> which is where those halt patches came from, but haven't finished.
>>
> 
> But what is the use case of semihosting on m68k?
> Is there a kernel implementation available for that?

Not that I'm aware of.  Do you want to remove it entirely instead?
Leaving it as-is should not be an option...


r~

Le 08/06/2022 à 21:05, Richard Henderson a écrit :
> On 6/8/22 11:08, Laurent Vivier wrote:
>> Le 08/06/2022 à 16:53, Richard Henderson a écrit :
>>> On 6/8/22 02:36, Laurent Vivier wrote:
>>>> Le 08/06/2022 à 00:26, Richard Henderson a écrit :
>>>>> Based-on: <20220607204557.658541-1-richard.henderson@linaro.org>
>>>>> ("[PATCH v4 00/53] semihosting cleanup")
>>>>>
>>>>> Changes for v4:
>>>>>    * Split out of v2.
>>>>>    * Convert host errno to gdb errno, which for m68k is guest errno.
>>>>>
>>>>
>>>> How do you test semihosting on m68k?
>>>
>>> I have only compiled this.  I had been working on adding something to tests/tcg/m68k/, which is 
>>> where those halt patches came from, but haven't finished.
>>>
>>
>> But what is the use case of semihosting on m68k?
>> Is there a kernel implementation available for that?
> 
> Not that I'm aware of.  Do you want to remove it entirely instead?

No, while you fix it I have no problem to keep it.

Thanks,
Laurent

On 6/8/22 12:05, Richard Henderson wrote:
>> But what is the use case of semihosting on m68k?
>> Is there a kernel implementation available for that?
> 
> Not that I'm aware of.  Do you want to remove it entirely instead?
> Leaving it as-is should not be an option...

Oh, duh.  The "kernel" implementation is libgloss itself.
The use-case is running the gcc testsuite for coldfire under qemu.


r~