On 01/25/2017 04:28 PM, Yao Qi wrote:
> On 17-01-25 12:11:01, Luis Machado wrote:
>> That is a reasonable assessment. insn-reverse.[c|exp] is redundant and IMO
>> would benefit from renaming too.
>>
>> The support in "insn-support-<arch>.c means support for a set of
>> instructions for this particular subsystem of gdb, therefore why i went with
>> that name. Thinking about it further, instruction decoding support is the
>> basis/foundation of reverse debugging, without which things would not work
>> properly. But i may be overthinking. :-)
>
> Every test is about testing some sort of support.  Breakpoint test is
> about breakpoint support, tracepoint test is about tracepoint support.
> We don't have to explicitly mention "support" in the test case name,
> IMO.
>
> It is easy to relate "insn-reverse-<arch>.c" to "insn-reverse.c".
> If you think "reverse" is redundant, "insn.c" and "insn-<arch>.c" is
> acceptable to me too.
>

It is not terribly important. I've reverted to the original name 
(insn-reverse-<arch>.c), updated things to mention the new name and 
pushed this as 8b00c176168dc7b0d78d0dc1f7d42f915375dc4a.

Patch attached.

Thanks for reviewing,
Luis