On 14-05-2020 17:32, Tom Tromey wrote:
>>>>>> "Tom" == Tom de Vries <tdevries@suse.de> writes:
> 
> Tom> Fix this by turning this and similar test-cases into regular, single
> Tom> executable test-cases.
> 
> Tom> This fixes 100+ FAILs with target board native-gdbserver.
> 
> I don't have any issue with splitting, however I wonder...
> 
> Tom> 	* gdb.base/align.exp.in: ...
> 
> ... if we should have a different suffix than ".in".
> To me ".in" normally implies that it's processed by configure or the like.
> 

FWIW, I based this pattern on:
...
$ ls -1  gdb/testsuite/gdb.base/all-architecture*
gdb/testsuite/gdb.base/all-architectures-0.exp
gdb/testsuite/gdb.base/all-architectures-1.exp
gdb/testsuite/gdb.base/all-architectures-2.exp
gdb/testsuite/gdb.base/all-architectures-3.exp
gdb/testsuite/gdb.base/all-architectures-4.exp
gdb/testsuite/gdb.base/all-architectures-5.exp
gdb/testsuite/gdb.base/all-architectures-6.exp
gdb/testsuite/gdb.base/all-architectures-7.exp
gdb/testsuite/gdb.base/all-architectures.exp.in
...

I'm guessing .in was short for include.

> Could we just call these ".tcl" files?  If dejagnu would ignore those,
> then that would work out great.

Testing attached patch, renaming *.exp.in to *.exp.tcl.

That leaves just:
...
$ find gdb/testsuite/gdb.* -type f -name "*.in"
gdb/testsuite/gdb.guile/scm-objfile-script-gdb.in
gdb/testsuite/gdb.python/py-framefilter-invalidarg-gdb.py.in
gdb/testsuite/gdb.python/py-framefilter-gdb.py.in
...

Thanks,
- Tom