5.1-56-ga6cf24b Crashes when pasting nothing.

[waxhead]

I noticed RT crashing a while back, like last week, but couldn't work out what triggered the crash, but I knew it was around copying and pasting edits from one image to another. Subsequent attempts to replicate the crash failed.

Tonight I did it again, but figured the issue was partly due to me missing the C in the CRTL-C operation and then hitting CRTL-V to paste nothing.

From the Browser windows open an image to the Editor window, click on it in the film strip hit CRTL-V, without CRTL-C. This should work. An alternative approach is to click on an image that has nothing down to it and CRTL-C CRTL-V that on your working image.

I've compiled a debug version of the latest from dev tonight and captured the debug log, attached. Info from about:

Version: 5.1-56-ga6cf24b
Branch: dev
Commit: a6cf24b
Commit date: 2017-06-12
Compiler: cc 5.4.0
Processor: x86_64
System: Linux
Bit depth: 64 bits
Gtkmm: V3.18.0
Build type: debug
Build flags: -std=c++11 -march=native -Werror=unused-label -fopenmp -Werror=unknown-pragmas -Wall -Wno-unused-result -Wno-deprecated-declarations -g
Link flags: -march=native
OpenMP support: ON
MMAP support: ON

rt-crash-crtl-v-no-crtl-c.txt

[heckflosse]

@waxhead Thanks for the report. I tried to reproduce it on Win7/64 and on Linux with exactly the same rt version you used. Could not reproduce it :( There must be something missing.

[waxhead]

@heckflosse I was just about to go to bed.

OK, so I opened up the non debug version of RT and I had to do it a couple of times to get it to actually crash.

I repeatedly did CRTL-V nothing, but then I copy and pasted from an image that has no changes against it, and it crashed.

I loaded up RT again, and this time I stayed in the file browser and did a CRTL-C and CRTL-V from one image to another, did a second CRTL-V and it crashed.

Sometimes I get a core dumped other times I get a lot of output on the console.

This is on linux too.

[heckflosse]

@waxhead
Ok, I will try harder

[waxhead]

I'm attaching the console output from the dev build, not the debug build.

console-output.txt

[waxhead]

OK I can do this more reliably now.

Copy CRTL-C from an image. CRTL-V to another image, CRTL-V again, Crash.

[heckflosse]

Bug confirmed

[waxhead]

From a usability point of view I was editing a couple of images and I copies the settings from my first image, pasted to a second one, didn't like the result and undid it. Opened a different image and just wanted to apply the setting again, second CRTL-V crashed RT. So it might appear that it's any additional CRTL-V after a the first one, even on different images.

This is what was in the console:

pete@kanga~/rt $ ./rawtherapee

(rawtherapee:18881): glibmm-ERROR **:
unhandled exception (type std::exception) in signal handler:
what: vector::_M_range_check: __n (which is 0) >= this->size() (which is 0)

Trace/breakpoint trap (core dumped)

Probably not that helpful...