scan-build

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
4 messages Options
Reply | Threaded
Open this post in threaded view
|

scan-build

Paul Ramsey
I see “scan-build” has entered the CI zoo on master, and I see it has borked a PR, but I cannot see *why* it has borked it…

Here’s the PR

https://github.com/libgeos/geos/pull/252

Here’s the build

https://travis-ci.com/libgeos/geos/jobs/266448653

This is not actionable information, how is someone who has tripped this wire supposed to respond?

P.
_______________________________________________
geos-devel mailing list
[hidden email]
https://lists.osgeo.org/mailman/listinfo/geos-devel
Reply | Threaded
Open this post in threaded view
|

Re: scan-build

Daniel Baston
Yeah, the reporting is pretty bad. If you search the log for "warning" you will find the cause:


which is a false positive. The build shows as failed since it's from 3 months ago; Raul has since changed the configuration so that scan-build reports but does not fail the build.

Dan

On Mon, Mar 9, 2020 at 6:20 PM Paul Ramsey <[hidden email]> wrote:
I see “scan-build” has entered the CI zoo on master, and I see it has borked a PR, but I cannot see *why* it has borked it…

Here’s the PR

https://github.com/libgeos/geos/pull/252

Here’s the build

https://travis-ci.com/libgeos/geos/jobs/266448653

This is not actionable information, how is someone who has tripped this wire supposed to respond?

P.
_______________________________________________
geos-devel mailing list
[hidden email]
https://lists.osgeo.org/mailman/listinfo/geos-devel

_______________________________________________
geos-devel mailing list
[hidden email]
https://lists.osgeo.org/mailman/listinfo/geos-devel
Reply | Threaded
Open this post in threaded view
|

Re: scan-build

rmrodriguez
Yes, what Dan said is true.

After updating clang that false positive is gone and the build is
clean, but now a failure is accepted. Depending on useful it turns out
to be I might do some grep-foo to just throw the errors and summary in
the future.

On Tue, Mar 10, 2020 at 2:35 AM Daniel Baston <[hidden email]> wrote:

>
> Yeah, the reporting is pretty bad. If you search the log for "warning" you will find the cause:
>
> https://travis-ci.com/libgeos/geos/jobs/266448653#L1082
>
> which is a false positive. The build shows as failed since it's from 3 months ago; Raul has since changed the configuration so that scan-build reports but does not fail the build.
>
> Dan
>
> On Mon, Mar 9, 2020 at 6:20 PM Paul Ramsey <[hidden email]> wrote:
>>
>> I see “scan-build” has entered the CI zoo on master, and I see it has borked a PR, but I cannot see *why* it has borked it…
>>
>> Here’s the PR
>>
>> https://github.com/libgeos/geos/pull/252
>>
>> Here’s the build
>>
>> https://travis-ci.com/libgeos/geos/jobs/266448653
>>
>> This is not actionable information, how is someone who has tripped this wire supposed to respond?
>>
>> P.
>> _______________________________________________
>> geos-devel mailing list
>> [hidden email]
>> https://lists.osgeo.org/mailman/listinfo/geos-devel
>
> _______________________________________________
> geos-devel mailing list
> [hidden email]
> https://lists.osgeo.org/mailman/listinfo/geos-devel



--
Raúl Marín Rodríguez
carto.com
_______________________________________________
geos-devel mailing list
[hidden email]
https://lists.osgeo.org/mailman/listinfo/geos-devel
Reply | Threaded
Open this post in threaded view
|

Re: scan-build

Even Rouault-2
Hi,

> Yes, what Dan said is true.
>
> After updating clang that false positive is gone and the build is
> clean, but now a failure is accepted. Depending on useful it turns out
> to be I might do some grep-foo to just throw the errors and summary in
> the future.

I faced a similar issue with GDAL.

The following might help to create a human friendly summary:

- in https://github.com/OSGeo/gdal/blob/master/gdal/ci/travis/csa_part_1/install.sh#L14
       
   ==> Using -o to output to a controled directory and -sarif for the json-based output format

- https://github.com/OSGeo/gdal/blob/master/gdal/ci/travis/csa_common/script.sh

   To postprocess all files written in the output directory and give first a detailed
   output, but not necessarily easy to understand. And a final summary section that contains
   just the filename, line and warning message.
   You may need to edit line 10 to remove a few GDAL specific exclusions of false
   positives (or things in external code we don't/wan't to control)

Even

--
Spatialys - Geospatial professional services
http://www.spatialys.com
_______________________________________________
geos-devel mailing list
[hidden email]
https://lists.osgeo.org/mailman/listinfo/geos-devel