Re: [PATCH] Add minimal Hexagon target - First in a series of patches - linux-user changes + linux-user/hexagon + skeleton of target/hexagon - Files in target/hexagon/imported are from another project and therefore do not conform to qemu coding standards

["Alex Bennée" <alex.bennee@linaro.org>]

Taylor Simpson <tsimpson@quicinc.com> writes:

> Is there a precedent for this?  I'm OK with DEBUG_HEX, but I assumed reviewers wouldn't approve
> #ifdef FIXME
> #define DEBUG_HEX
> #endif

We are trying to phase out this style of #define as it's prone to
bitrot. What you can do if define a debug helper like:

#define D(fmt, ..) \
  if (DEBUG_HEXAGON) { \
      qemu_log(fmt, ## __VA_ARGS__); \
  }

So the debug will still be built (and format strings checked etc) but
with DEBUG_HEXAGON as 0 things get dead coded away.

>
> Taylor
>
>
> -----Original Message-----
> From: Richard Henderson <richard.henderson@linaro.org>
> Sent: Wednesday, November 20, 2019 3:02 AM
> To: Laurent Vivier <laurent@vivier.eu>; Taylor Simpson <tsimpson@quicinc.com>; Philippe Mathieu-Daudé <philmd@redhat.com>; riku.voipio@iki.fi; qemu-devel@nongnu.org
> Subject: Re: [PATCH] Add minimal Hexagon target - First in a series of patches - linux-user changes + linux-user/hexagon + skeleton of target/hexagon - Files in target/hexagon/imported are from another project and therefore do not conform to qemu coding standards
>
>
> On 11/20/19 9:33 AM, Laurent Vivier wrote:
>> Le 20/11/2019 à 05:48, Taylor Simpson a écrit :
>>> For the general DEBUG_HEX messages, I think the trace infrastructure isn't quite what I'm looking for.
>>>
>>> Here's a sample of what it prints
>>> Start packet: pc = 0x4002f0
>>> Packet committed: pc = 0x4002f0
>>> Regs written
>>>         r18 = -69420 (0xfffef0d4)
>>>         r19 = -69412 (0xfffef0dc)
>>> Stores
>>>         memd[0xfffef0b0] = 0 (0x0000000000000000) Next PC = 0x4002f8
>>> Exec counters: pkt = 17, insn = 24, hvx = 0
>
> For something like this, I'd keep DEBUG_HEX.
>
>>>     if (qemu_loglevel_maks(CPU_LOG_EXEC)) {
>
> CPU_LOG_EXEC already has a specific structure, listing the TranslationBlocks that are executed.  It shouldn't be hijacked for something else.
>
> If you really want a runtime flag for this, we should add a new CPU_LOG_* flag.
>
>
> r~


--
Alex Bennée