[PATCH 0/5] Solve caching problems with compiler_info

[Andrew Burgess]

Tom de Vries via Gdb-patches <gdb-patches@sourceware.org> writes:

> On 6/9/22 15:42, Andrew Burgess via Gdb-patches wrote:
>> "Kempke, Nils-Christian via Gdb-patches" <gdb-patches@sourceware.org>
>> writes:
>> 
>>>> -----Original Message-----
>>>> From: Gdb-patches <gdb-patches-bounces+nils-
>>>> christian.kempke=intel.com@sourceware.org> On Behalf Of Andrew
>>>> Burgess via Gdb-patches
>>>> Sent: Wednesday, June 8, 2022 4:22 PM
>>>> To: gdb-patches@sourceware.org
>>>> Cc: Andrew Burgess <aburgess@redhat.com>
>>>> Subject: [PATCH 0/5] Solve caching problems with compiler_info
>>>>
>>>> This series is an alternative proposal to solve the caching problems
>>>> with compiler_info.
>>>>
>>>> Patch #1 is taken unmodified from Nils' v2 series.
>>>>
>>>> Patches #2, #3, #4, and #5 are new work, though the final patch
>>>> addresses the same problem as the second patch in Nils' v2 series.
>>>>
>>>> Thanks,
>>>> Andrew
>>>
>>>
>>> Hi Andrew,
>>>
>>> Thanks for this - I only had two minor comments really. The rest of this
>>> series is nice.  I do prefer it over my initially submitted patches. And thanks
>>> for showing how the associative array works !
>> 
>> Thanks for the feedback.
>> 
>> As this issue was causing significant test failures when not running
>> tests in parallel, I've gone ahead and pushed this (with the updated
>> patch #2).
>> 
>> I'm happy to address any feedback that comes up later.
>
> I'm running into:
> ...
> ERROR: tcl error sourcing 
> /home/vries/gdb_versions/devel/src/gdb/testsuite/gdb.base/all-bin.exp.
> ERROR: can't read "false": no such variable
>      while executing
> "gdb_test "print v_int <= v_short" " = $false" \
>      "print value of v_int<=v_short""
> ...
> which doesn't occur if I run the test-case in isolation.
>
> Currently trying out:
> ...
> diff --git a/gdb/testsuite/lib/gdb.exp b/gdb/testsuite/lib/gdb.exp
> index 67e5838a7a7..37a28822311 100644
> --- a/gdb/testsuite/lib/gdb.exp
> +++ b/gdb/testsuite/lib/gdb.exp
> @@ -4193,8 +4193,9 @@ proc get_compiler_info {{language "c"}} {
>
>       # Most compilers will evaluate comparisons and other boolean
>       # operations to 0 or 1.
> -    uplevel \#0 { set true 1 }
> -    uplevel \#0 { set false 0 }
> +    gdb_persistent_global true false
> +    set true 1
> +    set false 0
>
>       return 0
>   }
> ...
> which got past that test-case without running into the error.

Sorry for that.  I pushed the fix from:

  https://sourceware.org/pipermail/gdb-patches/2022-June/189985.html

after your review.

Thanks,
Andrew