Hi Tom
The tests were wrong, not the code. This change was made (by Bob) in
1997:
* lib/target.exp(set_target_info, compile, archive, ranlib,
link_objects, execute_anywhere, getprompt, make): Deleted.
(push_target,push_config,pop_config,pop_target): Don't copy
the
On Sat, Jun 03, 2017 at 07:07:03PM +0200, Tom de Vries wrote:
> The current failure in target.test is:
> ...
> can't read "target_info(target,ldflags)": no such element in array
> while executing
Can you please send a patch in the first instance that just produces a
proper FAIL rather than an
On Sat, Jun 03, 2017 at 06:57:08PM +0200, Tom de Vries wrote:
> invalid command name "rlogin"
> while executing
> "rlogin foobar.barfoo.com"
> invoked from within
> "if { [rlogin foobar.barfoo.com] < 0 } {
[...]
In the bad old days, telnet, rlogin, rsh, etc. were just procs. eg:
-
W
On Sat, Jun 03, 2017 at 06:33:37PM +0200, Tom de Vries wrote:
> this patch removes a test of proc list_targets in target.test. The proc was
> removed in this commit, so the test fails atm:
> Mon Jul 21 14:30:01 1997 Bob Manson
Thanks, Tom, for updating the testsuite 20 years after Bob's patch
On Sat, Jun 03, 2017 at 06:15:59PM +0200, Tom de Vries wrote:
> 2017-06-03 Tom de Vries
>
> * testsuite/runtest.all/remote.test (load_lib): New proc.
> * testsuite/runtest.all/target.test (load_lib): Same.
This, and the patch it depends on, are now in master.
Thanks,
Ben
signat
On Sat, Jun 03, 2017 at 05:59:52PM +0200, Tom de Vries wrote:
> 2017-06-03 Tom de Vries
>
> * testsuite/runtest.all/libs.exp (process_test): Check if test has
> reportable output.
Good catch, thanks, Tom!
Cheers,
Ben
signature.asc
Description: Digital signature
The current failure in target.test is:
...
can't read "target_info(target,ldflags)": no such element in array
while executing
"if { $target_info(target,ldflags) == "-Tidp.ld" } {
incr matches
}"
(file "../src/testsuite/runtest.all/target.test" line 76)
...
The test fails here:
...
pus
Hi,
The current failure in remote.test is:
-
Running ../src/testsuite/runtest.all/libs.exp ...
ERROR: ../src/testsuite/runtest.all/remote.test failed
../src/testsuite/runtest.all/remote.test stdout: ...
... done
../src/testsuite/runtest.all/remote.test
Hi,
this patch removes a test of proc list_targets in target.test. The proc
was removed in this commit, so the test fails atm:
...
Mon Jul 21 14:30:01 1997 Bob Manson
* lib/target.exp(list_targets): Deleted, no longer meaningful.
(default_target_compile): Use warning instead
[ Fail verbosely when test fails in process_test ]
On 06/03/2017 06:10 PM, Tom de Vries wrote:
[ was: Re: Check if test has reportable output in process_test ]
On 06/03/2017 05:59 PM, Tom de Vries wrote:
Hi,
I found a problem in the dejagnu testsuite, with process_test.
If there is a problem
[ was: Re: Check if test has reportable output in process_test ]
On 06/03/2017 05:59 PM, Tom de Vries wrote:
Hi,
I found a problem in the dejagnu testsuite, with process_test.
If there is a problem running a .test file, then process_test is silent
about it.
This minimal demonstrator patch a
Hi,
I found a problem in the dejagnu testsuite, with process_test.
If there is a problem running a .test file, then process_test is silent
about it.
This minimal demonstrator patch adds basic reporting if there is a
problem, by checking that at least a single clause in the expect command
tr
On 06/01/2017 06:27 PM, Rob Savoye wrote:
On 06/01/2017 10:23 AM, Tom de Vries wrote:
I have only tested this with the local testsuite.
I can test this more extensively with the gcc testsuite. I suppose I'll
have to test both:
The true validation test would be a full run of cross-testing
13 matches
Mail list logo
