[v2] send-error-report: improve debugging

From: Thomas Perrot <thomas.perrot@bootlin.com>

- add a debug mode
- print the request and the response when an error occurs.

Signed-off-by: Thomas Perrot <thomas.perrot@bootlin.com>
---

Changes in version 2:
- Drop the redundant checks

scripts/send-error-report | 25 ++++++++++++++++++++-----
 1 file changed, 20 insertions(+), 5 deletions(-)

--
2.49.0

Hi Thomas,

On 4/18/25 3:35 PM, Thomas Perrot via lists.openembedded.org wrote:
> From: Thomas Perrot <thomas.perrot@bootlin.com>
> 
> - add a debug mode
> - print the request and the response when an error occurs.
> 
> Signed-off-by: Thomas Perrot <thomas.perrot@bootlin.com>

Reviewed-by: Quentin Schulz <quentin.schulz@cherry.de>

Thanks!
Quentin

I noticed that this (or at least v1) was applied on master - would like
to call some attention to a related discussion[1] in a neighboring list,
and some patches[2][3][4] that followed it - at least [3] is a
pre-requisite for this patch (error-report-web has hardcoded the version
of the script to 0.3).

[1]: https://lists.yoctoproject.org/g/yocto/message/65184
[2]: https://lists.yoctoproject.org/g/yocto/message/65188
[3]: https://lists.yoctoproject.org/g/yocto/message/65199
[4]: https://lists.yoctoproject.org/g/yocto/message/65200

On 4/18/25 15:35, Thomas Perrot via lists.openembedded.org wrote:
> From: Thomas Perrot <thomas.perrot@bootlin.com>
>
> - add a debug mode
> - print the request and the response when an error occurs.
>
> Signed-off-by: Thomas Perrot <thomas.perrot@bootlin.com>
> ---
>
> Changes in version 2:
> - Drop the redundant checks
>
> scripts/send-error-report | 25 ++++++++++++++++++++-----
>  1 file changed, 20 insertions(+), 5 deletions(-)
>
> diff --git a/scripts/send-error-report b/scripts/send-error-report
> index cfbcaa52cbc3..9ea7160f52c4 100755
> --- a/scripts/send-error-report
> +++ b/scripts/send-error-report
> @@ -6,6 +6,7 @@
>  # Copyright (C) 2013 Intel Corporation
>  # Author: Andreea Proca <andreea.b.proca@intel.com>
>  # Author: Michael Wood <michael.g.wood@intel.com>
> +# Author: Thomas Perrot <thomas.perrot@bootlin.com>
>  #
>  # SPDX-License-Identifier: GPL-2.0-only
>  #
> @@ -22,7 +23,7 @@ scripts_lib_path = os.path.join(os.path.dirname(os.path.realpath(__file__)), 'li
>  sys.path.insert(0, scripts_lib_path)
>  import argparse_oe
>
> -version = "0.3"
> +version = "0.4"
>
>  log = logging.getLogger("send-error-report")
>  logging.basicConfig(format='%(levelname)s: %(message)s')
> @@ -140,14 +141,21 @@ def send_data(data, args):
>          url = args.protocol+args.server+"/ClientPost/"
>
>      req = urllib.request.Request(url, data=data, headers=headers)
> +
> +    log.debug(f"Request URL: {url}")
> +    log.debug(f"Request Headers: {headers}")
> +    log.debug(f"Request Data: {data.decode('utf-8')}")
> +
>      try:
>          response = urllib.request.urlopen(req)
>      except urllib.error.HTTPError as e:
> -        logging.error(str(e))
> +        log.error(f"HTTP Error {e.code}: {e.reason}")
> +        log.debug(f"Response Content: {e.read().decode('utf-8')}")
>          sys.exit(1)
>
> -    print(response.read().decode('utf-8'))
> -
> +    log.debug(f"Response Status: {response.status}")
> +    log.debug(f"Response Headers: {response.getheaders()}")
> +    log.info(response.read().decode('utf-8'))
>
>  if __name__ == '__main__':
>      arg_parse = argparse_oe.ArgumentParser(description="This scripts will send an error report to your specified error-report-web server.")
> @@ -195,13 +203,20 @@ if __name__ == '__main__':
>                             dest="protocol",
>                             action="store_const", const="http://", default="https://")
>
> -
> +    arg_parse.add_argument("-d",
> +                           "--debug",
> +                           help="Enable debug mode to print request/response details",
> +                           action="store_true")
>
>      args = arg_parse.parse_args()
>
>      if (args.json == False):
>          print("Preparing to send errors to: "+args.server)
>
> +    # Enable debugging if requested
> +    if args.debug:
> +        log.setLevel(logging.DEBUG)
> +
>      data = prepare_data(args)
>      send_data(data, args)
>
> --
> 2.49.0
>
> -=-=-=-=-=-=-=-=-=-=-=-
> Links: You receive all messages sent to this group.
> View/Reply Online (#215131): https://lists.openembedded.org/g/openembedded-core/message/215131
> Mute This Topic: https://lists.openembedded.org/mt/112330994/6084445
> Group Owner: openembedded-core+owner@lists.openembedded.org
> Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub [skandigraun@gmail.com]
> -=-=-=-=-=-=-=-=-=-=-=-
>

Hello,

On Sun, 2025-04-20 at 12:44 +0200, Gyorgy Sarvari via
lists.openembedded.org wrote:
> I noticed that this (or at least v1) was applied on master - would
> like
> to call some attention to a related discussion[1] in a neighboring
> list,
> and some patches[2][3][4] that followed it - at least [3] is a
> pre-requisite for this patch (error-report-web has hardcoded the
> version
> of the script to 0.3).
> 
> [1]: https://lists.yoctoproject.org/g/yocto/message/65184
> [2]: https://lists.yoctoproject.org/g/yocto/message/65188
> [3]: https://lists.yoctoproject.org/g/yocto/message/65199
> [4]: https://lists.yoctoproject.org/g/yocto/message/65200
> 

Can you revert the patch v1 and then apply patch v2, at least after
patch [3] has been applied to error-report-web?

Kind regards,
Thomas Perrot

> On 4/18/25 15:35, Thomas Perrot via lists.openembedded.org wrote:
> > From: Thomas Perrot <thomas.perrot@bootlin.com>
> > 
> > - add a debug mode
> > - print the request and the response when an error occurs.
> > 
> > Signed-off-by: Thomas Perrot <thomas.perrot@bootlin.com>
> > ---
> > 
> > Changes in version 2:
> > - Drop the redundant checks
> > 
> > scripts/send-error-report | 25 ++++++++++++++++++++-----
> >  1 file changed, 20 insertions(+), 5 deletions(-)
> > 
> > diff --git a/scripts/send-error-report b/scripts/send-error-report
> > index cfbcaa52cbc3..9ea7160f52c4 100755
> > --- a/scripts/send-error-report
> > +++ b/scripts/send-error-report
> > @@ -6,6 +6,7 @@
> >  # Copyright (C) 2013 Intel Corporation
> >  # Author: Andreea Proca <andreea.b.proca@intel.com>
> >  # Author: Michael Wood <michael.g.wood@intel.com>
> > +# Author: Thomas Perrot <thomas.perrot@bootlin.com>
> >  #
> >  # SPDX-License-Identifier: GPL-2.0-only
> >  #
> > @@ -22,7 +23,7 @@ scripts_lib_path =
> > os.path.join(os.path.dirname(os.path.realpath(__file__)), 'li
> >  sys.path.insert(0, scripts_lib_path)
> >  import argparse_oe
> > 
> > -version = "0.3"
> > +version = "0.4"
> > 
> >  log = logging.getLogger("send-error-report")
> >  logging.basicConfig(format='%(levelname)s: %(message)s')
> > @@ -140,14 +141,21 @@ def send_data(data, args):
> >          url = args.protocol+args.server+"/ClientPost/"
> > 
> >      req = urllib.request.Request(url, data=data, headers=headers)
> > +
> > +    log.debug(f"Request URL: {url}")
> > +    log.debug(f"Request Headers: {headers}")
> > +    log.debug(f"Request Data: {data.decode('utf-8')}")
> > +
> >      try:
> >          response = urllib.request.urlopen(req)
> >      except urllib.error.HTTPError as e:
> > -        logging.error(str(e))
> > +        log.error(f"HTTP Error {e.code}: {e.reason}")
> > +        log.debug(f"Response Content: {e.read().decode('utf-8')}")
> >          sys.exit(1)
> > 
> > -    print(response.read().decode('utf-8'))
> > -
> > +    log.debug(f"Response Status: {response.status}")
> > +    log.debug(f"Response Headers: {response.getheaders()}")
> > +    log.info(response.read().decode('utf-8'))
> > 
> >  if __name__ == '__main__':
> >      arg_parse = argparse_oe.ArgumentParser(description="This
> > scripts will send an error report to your specified error-report-
> > web server.")
> > @@ -195,13 +203,20 @@ if __name__ == '__main__':
> >                             dest="protocol",
> >                             action="store_const", const="http://",
> > default="https://")
> > 
> > -
> > +    arg_parse.add_argument("-d",
> > +                           "--debug",
> > +                           help="Enable debug mode to print
> > request/response details",
> > +                           action="store_true")
> > 
> >      args = arg_parse.parse_args()
> > 
> >      if (args.json == False):
> >          print("Preparing to send errors to: "+args.server)
> > 
> > +    # Enable debugging if requested
> > +    if args.debug:
> > +        log.setLevel(logging.DEBUG)
> > +
> >      data = prepare_data(args)
> >      send_data(data, args)
> > 
> > --
> > 2.49.0
> > 
> > 
> > 
> 
> 
> -=-=-=-=-=-=-=-=-=-=-=-
> Links: You receive all messages sent to this group.
> View/Reply Online (#215153):
> https://lists.openembedded.org/g/openembedded-core/message/215153
> Mute This Topic: https://lists.openembedded.org/mt/112330994/5443093
> Group Owner: openembedded-core+owner@lists.openembedded.org
> Unsubscribe:
> https://lists.openembedded.org/g/openembedded-core/unsub [
> thomas.perrot@bootlin.com]
> -=-=-=-=-=-=-=-=-=-=-=-
>

On Fri, 2025-04-25 at 17:02 +0200, Thomas Perrot via
lists.openembedded.org wrote:
> Hello,
> 
> On Sun, 2025-04-20 at 12:44 +0200, Gyorgy Sarvari via
> lists.openembedded.org wrote:
> > I noticed that this (or at least v1) was applied on master - would
> > like
> > to call some attention to a related discussion[1] in a neighboring
> > list,
> > and some patches[2][3][4] that followed it - at least [3] is a
> > pre-requisite for this patch (error-report-web has hardcoded the
> > version
> > of the script to 0.3).
> > 
> > [1]: https://lists.yoctoproject.org/g/yocto/message/65184
> > [2]: https://lists.yoctoproject.org/g/yocto/message/65188
> > [3]: https://lists.yoctoproject.org/g/yocto/message/65199
> > [4]: https://lists.yoctoproject.org/g/yocto/message/65200
> > 
> 
> Can you revert the patch v1 and then apply patch v2, at least after
> patch [3] has been applied to error-report-web?

It would be helpful to address that to the person who maintains the
repo. I've updated to the v2 of your patch.

I've queued your patch for the server but updating the instance is
beyond my access/skills. I also queued a django and python3 upgrade on
master-next there which it would be good to test.

Copying helpdesk to see if they can schedule in some help in updating
the instance either to master (easier) or ultimately to test master-
next and move to that.

Cheers,

Richard

The new Django 4.2.20 https://errors.yoctoproject.org/ is live. The
previous version is still available at
https://errors-prev.yoctoproject.org/Errors/Latest/Autobuilder/ for
comparison. Tracking at
https://bugzilla.yoctoproject.org/show_bug.cgi?id=15336.


Thanks to everyone who helped with the upgrade.


On Fri, Apr 25, 2025 at 8:19 AM Richard Purdie via lists.openembedded.org
<richard.purdie=linuxfoundation.org@lists.openembedded.org> wrote:

> On Fri, 2025-04-25 at 17:02 +0200, Thomas Perrot via
> lists.openembedded.org wrote:
> > Hello,
> >
> > On Sun, 2025-04-20 at 12:44 +0200, Gyorgy Sarvari via
> > lists.openembedded.org wrote:
> > > I noticed that this (or at least v1) was applied on master - would
> > > like
> > > to call some attention to a related discussion[1] in a neighboring
> > > list,
> > > and some patches[2][3][4] that followed it - at least [3] is a
> > > pre-requisite for this patch (error-report-web has hardcoded the
> > > version
> > > of the script to 0.3).
> > >
> > > [1]: https://lists.yoctoproject.org/g/yocto/message/65184
> > > [2]: https://lists.yoctoproject.org/g/yocto/message/65188
> > > [3]: https://lists.yoctoproject.org/g/yocto/message/65199
> > > [4]: https://lists.yoctoproject.org/g/yocto/message/65200
> > >
> >
> > Can you revert the patch v1 and then apply patch v2, at least after
> > patch [3] has been applied to error-report-web?
>
> It would be helpful to address that to the person who maintains the
> repo. I've updated to the v2 of your patch.
>
> I've queued your patch for the server but updating the instance is
> beyond my access/skills. I also queued a django and python3 upgrade on
> master-next there which it would be good to test.
>
> Copying helpdesk to see if they can schedule in some help in updating
> the instance either to master (easier) or ultimately to test master-
> next and move to that.
>
> Cheers,
>
> Richard
>
> -=-=-=-=-=-=-=-=-=-=-=-
> Links: You receive all messages sent to this group.
> View/Reply Online (#215480):
> https://lists.openembedded.org/g/openembedded-core/message/215480
> Mute This Topic: https://lists.openembedded.org/mt/112330994/1003190
> Group Owner: openembedded-core+owner@lists.openembedded.org
> Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub [
> mhalstead@linuxfoundation.org]
> -=-=-=-=-=-=-=-=-=-=-=-
>
>