Re: pvAccess assert error

["Johnson, Andrew N. via Tech-talk" <tech-talk at aps.anl.gov>]

Hi Mark,

 

On 2/29/24, 1:08 PM, "Tech-talk" tech-talk-bounces at aps.anl.gov wrote:

 

It would have saved time and frustration if this bug and fix had been contained in the release notes for base 7.0.8 or pvAccessCPP 7.1.7.

 

Sorry, you can blame me for that although other than the code changes the only information I had about the fix when I updated the Release notes for pvAccessCPP 7.1.7 were these two git log messages from Freddie:

 

d4bb5952a Fix crash when toEcho empty

55ab0fd3b Avoid data() as it is c++11

 

They both came from this PR which doesn't give much more explanation.

 

- Andrew

 

-- 

Complexity comes for free, Simplicity you have to work for.

 

 

 

 

 

 

 

-----Original Message-----
From: Michael Davidsaver <mdavidsaver at gmail.com>
Sent: Wednesday, February 28, 2024 4:22 PM
To: Mark Rivers <rivers at cars.uchicago.edu>
Cc: tech-talk at aps.anl.gov
Subject: Re: pvAccess assert error

 

On 2/28/24 10:49, Mark Rivers wrote:

> Hi Michael,

>

> I sent a message to both tech-talk and core-talk at 4:31 PM CST yesterday.

>

> This is the tech-talk message:

> https://epics.anl.gov/tech-talk/2024/msg00292.php

>

> This is the core-talk message:

> https://epics.anl.gov/core-talk/2024/msg00042.php

 

Ah, this is an issue with MSVC debug builds which Freddie Akeroyd already found and fixed.

 

This fix is included in pvAccessCPP 7.1.7, which is bundled in epics-base 7.0.8.

 

 

> However, I did not actually then receive the message from tech-talk.  And it seems like maybe you did not either?  It seems like tech-talk archived it but did not send it?

>

> Mark

>

>

> -----Original Message-----

> From: Michael Davidsaver <mdavidsaver at gmail.com>

> Sent: Wednesday, February 28, 2024 12:31 PM

> To: Mark Rivers <rivers at cars.uchicago.edu>

> Subject: Re: pvAccess assert error

>

> Hi Mark,

>

> Am I being dense today?  I only see a screen capture attached to your original mail.

>

>

> On 2/28/24 10:25, Mark Rivers wrote:

>> Hi Michael,

>> 

>> Thanks for the reply.  I sent the full stack trace yesterday.

>> 

>> Mark

>> 

>> 

>> -----Original Message-----

>> From: Michael Davidsaver <mdavidsaver at gmail.com>

>> Sent: Wednesday, February 28, 2024 12:20 PM

>> To: Mark Rivers <rivers at cars.uchicago.edu>

>> Cc: tech-talk at aps.anl.gov

>> Subject: Re: pvAccess assert error

>> 

>> On 2/26/24 16:37, Mark Rivers via Tech-talk wrote:

>>> Folks,

>>> 

>>> I am running a new areaDetector camera.  It generates frames that are 4096x3072 pixels, 8-bit.  I can collect data from the camera at 150 frames/s using a BitFlow CoaXSpress card with no problems.  This is on a Windows PC.

>>> 

>>> However, when I try to use pvAccess to read the data from the NDPluginPva plugin I get an assert error in pvAccess in my Windows application.  The error message is in the attachment.

>>> 

>>> This happens when I use the ImageJ pvAccess plugin, and also when I use the following Linux command on another machine:

>>> 

>>> $ pvmonitor 13BF1:Pva1:Image > /dev/null

>>> 

>>> Any idea what the problem could be?  This is base 7.0.7.

>> 

>> At a guess, maybe another size()==0 situations as with:

>> 

>> https://github.com/epics-base/pvDataCPP/issues/50

>> 

>> If so, a full stack trace should point at the offending code.

>> 

>