Re: [PATCH] Implement nbsd_nat_target::thread_name

[Kamil Rytarowski <n54@gmx.com>]

On 01.04.2020 21:28, Tom Tromey wrote:
>>>>>> "Kamil" == Kamil Rytarowski <n54@gmx.com> writes:
>
> Kamil> Return the name assigned to a thread by an application.  Return
> Kamil> the string in a static buffer.
>
> Thanks for the patch.
>
> Kamil> +  if (sysctl (mib, 5, NULL, &size, NULL, 0) == -1 || size == 0)
> Kamil> +    perror_with_name (("sysctl"));
>
> I suspect it's better to return NULL there than to throw an exception.
>
> Kamil> +  if (kl == NULL)
> Kamil> +    perror_with_name (("calloc"));
> Kamil> +
> Kamil> +  if (sysctl (mib, 5, kl.get (), &size, NULL, 0) == -1 || size == 0)
> Kamil> +    perror_with_name (("sysctl"));
>
> Likewise in these cases.
>
> thanks,
> Tom
>

Please check a newer patch:

[PATCH v2] Implement basic threading support in the NetBSD target

https://sourceware.org/pipermail/gdb-patches/2020-March/167018.html

Does this comment still apply to return NULL for thread_name()
thread_alive() and update_thread_list()?

In practice the kernel will almost never return error when used
properly, unless there is specified a wrong process id or similar. It's
generally safe to abort on error here.