Re: [PATCH] gdb/testsuite/gdb.rocm: Fix incorrect use of continue N in multi-inferior-gpu.exp

[Simon Marchi <simark@simark.ca>]

On 10/17/23 10:41, Lancelot SIX wrote:
> 
>> On 10/17/23 07:16, Lancelot Six wrote:
>>> The gdb.rocm/multi-inferior-gpu.exp testcase uses a "continue $thread"
>>> command, but this is incorrect.  If "continue" is given an argument, it
>>> sets the ignore count of the breakpoint the thread stopped at.
>>>
>>> For this testcase it does not really matter since the breakpoint is not
>>> meant to be hit anymore, so whatever the ignore count is won't influence
>>> the outcome of the test.  It is worth fixing nevertheless.
>>>
>>> While at changing it, switch to using "continue&" to consume the GDB
>>> prompt right away.  This makes the following pattern matching more
>>> reliable.
>>
>> I don't understand the "more reliable" part, can you explain?
>>
>> Simon
> 
> There will be a "(gdb) " prompt appearing at some point after a "continue", but in non-stop we don't really control when it appears (if I understand correctly).  It seems that other non-stop tests use the "&" variant and consume the prompt immediately.
> 
> With this, before the next command is issued, we know that 1) we got a prompt and 2) we have seen the process we just released finish.
> 
> Overall, I am mostly trying to follow patterns I caught in other non-stop tests.

Ah, I missed that we were in non-stop.

If we run a single thread at a time, I think it will be reliable to use
the sync variant of continue.  Using the async variant is useful when
you resume multiple threads and expect multiple events (like breakpoint
hits) which can happen in any order.  At least that's my understanding.

Simon