Satellites: Restore shadow circle at arbitrary altitude #3538

gzotti · 2023-12-08T23:24:05Z

Description

The functionality to show a shadow limit for a hypothetical satellite in arbitrary altitude above ground was botched and therefore disabled in 23.1. This now aims to fix that.

also resets cutoff altitude to 80km
also no commLink info for nonoperational satellites.

As discussed: We had envisioned shadow enlargement like for Lunar eclipses. However, the original evaluation for visibility conditions does not use shadow enlargement. I have removed the according code.

Fixes #3100 (issue)

Screenshots (if appropriate):

Type of change

Bug fix (non-breaking change which fixes an issue)
New feature (non-breaking change which adds functionality)
Breaking change (fix or feature that would cause existing functionality to change)
This change requires a documentation update
Housekeeping

How Has This Been Tested?

Test Configuration:

Operating system: Win11
Graphics Card: Geforce 3700Ti

Checklist:

My code follows the code style of this project.
I have performed a self-review of my own code
I have commented my code, particularly in hard-to-understand areas
I have made corresponding changes to the documentation (header file)
I have updated the respective chapter in the Stellarium User Guide
My changes generate no new warnings
I have added tests that prove my fix is effective or that my feature works
New and existing unit tests pass locally with my changes
Any dependent changes have been merged and published in downstream modules

- also resets cutoff to 80km - also no commLink info for nonoperational satellites.

alex-w · 2023-12-09T03:13:23Z

However, the original evaluation for visibility conditions does not use shadow enlargement. Either remove the code, or add it to the original model?

I think the code for shadow enlargement should be at least disabled or removed. In the future we need improve brightness model for satellites in penumbra and we need a real observations to obtain info for accuracy of our umbra/penumbra for sats.

Stop! The planetarium is crashed at exit when umbra is visible.

gzotti · 2023-12-09T09:30:25Z

I see no crash at exit in about 10 runs. Can you be a bit more specific?

alex-w · 2023-12-09T09:47:43Z

I see no crash at exit in about 10 runs. Can you be a bit more specific?

Steps to reproduce:

Run Stellarium
Open Satellites settings and enable the umbra
Select ISS
Quit from Stellarium

Downloaded 0 files (0 kbytes) in a session of 49.06 sec (average of 0 kB/s + 0 files from cache (0 kB)).
LandscapeMgr: Clearing cache of 0 landscapes totalling about  0 MB.
StelGLWidget destroyed
[Thread 0x7fffd5476640 (LWP 55222) exited]
[Thread 0x7fffdd1ff640 (LWP 55144) exited]
[Thread 0x7fff5a111640 (LWP 55224) exited]
[Thread 0x7fff59910640 (LWP 55228) exited]
[Thread 0x7fffd4c75640 (LWP 55223) exited]
[Thread 0x7fffdc9fe640 (LWP 55145) exited]
[Thread 0x7fffe59fe640 (LWP 55140) exited]

Thread 1 "stellarium" received signal SIGSEGV, Segmentation fault.
0x0000000000000031 in ?? ()

gzotti · 2023-12-09T11:18:14Z

Still cannot reproduce it on W11. Stellarium on my WSL currently would use the Intel GPU but segfaults on startup. Due to other tasks I cannot reboot now. Can you run a Debug build and check where it breaks?

alex-w · 2023-12-09T11:21:22Z

Still cannot reproduce it on W11. Stellarium on my WSL currently would use the Intel GPU but segfaults on startup. Due to other tasks I cannot reboot now. Can you run a Debug build and check where it breaks?

I've shared log from gdb, so, the main problem - I see the segmentation fault, but I see no source. Something strange happened in new getUmbraData method.

alex-w · 2023-12-09T11:23:41Z

Oops... I've checked the master - same steps and same results - a crash O_O

So, your code is not a bad guy

10110111 · 2023-12-09T11:28:57Z

Thread 1 "stellarium" received signal SIGSEGV, Segmentation fault.
0x0000000000000031 in ?? ()

If it's 100% reproducible, try running under Valgrind. It may be able to tell where the bad jump/call target address came from.

alex-w · 2023-12-09T11:50:58Z

If it's 100% reproducible, try running under Valgrind. It may be able to tell where the bad jump/call target address came from.

Hmm...

master, debug mode, gdb:

Downloaded 0 files (0 kbytes) in a session of 34.195 sec (average of 0 kB/s + 0 files from cache (0 kB)).
LandscapeMgr: Clearing cache of 0 landscapes totalling about  0 MB.
StelGLWidget destroyed
[Thread 0x7fffdd413640 (LWP 122542) exited]
[Thread 0x7fffddc14640 (LWP 122541) exited]
[Thread 0x7fffdf298640 (LWP 122539) exited]
[Thread 0x7fffdcc12640 (LWP 122543) exited]
[Thread 0x7fff6f0fc640 (LWP 122566) exited]
[Thread 0x7fff700fe640 (LWP 122564) exited]
[Thread 0x7fff708ff640 (LWP 122563) exited]
[Thread 0x7fffdfba5ec0 (LWP 122534) exited]
[Thread 0x7fff6f8fd640 (LWP 122565) exited]
[New process 122534]
[Inferior 1 (process 122534) exited normally]

Am I an idiot?

OK, this branch, debug mode, gdb:

Downloaded 0 files (0 kbytes) in a session of 47.111 sec (average of 0 kB/s + 0 files from cache (0 kB)).
LandscapeMgr: Clearing cache of 0 landscapes totalling about  0 MB.
malloc(): unsorted double linked list corrupted

Thread 1 "stellarium" received signal SIGABRT, Aborted.
__pthread_kill_implementation (no_tid=0, signo=6, threadid=140736946921152) at ./nptl/pthread_kill.c:44
44      ./nptl/pthread_kill.c: Нет такого файла или каталога.

Next iteration:

LandscapeMgr: Clearing cache of 0 landscapes totalling about  0 MB.
StelGLWidget destroyed
[Thread 0x7fffddc14640 (LWP 135903) exited]
[Thread 0x7fffdd413640 (LWP 135904) exited]
[Thread 0x7fffdf298640 (LWP 135901) exited]
[Thread 0x7fffdcc12640 (LWP 135905) exited]
[Thread 0x7fff6f8fd640 (LWP 135931) exited]
[Thread 0x7fff700fe640 (LWP 135930) exited]
[Thread 0x7fff708ff640 (LWP 135929) exited]
[Thread 0x7fffdfba5ec0 (LWP 135897) exited]
[Thread 0x7fff6f0fc640 (LWP 135932) exited]
[New process 135897]
[Inferior 1 (process 135897) exited normally]

WTF?!?

10110111 · 2023-12-09T11:55:16Z

malloc(): unsorted double linked list corrupted

Well, this looks like that same intermittent crash that we've already seen and failed to diagnose.

alex-w · 2023-12-09T11:57:20Z

this branch, debug mode, gdb

Run Stellarium
Change time to 3 hours early (a daylight on the screen)
Open settings Satellites plugin
Umbra and fixed altitude enabled
Quit

Downloaded 0 files (0 kbytes) in a session of 51.29 sec (average of 0 kB/s + 0 files from cache (0 kB)).
LandscapeMgr: Clearing cache of 0 landscapes totalling about  0 MB.

Thread 1 "stellarium" received signal SIGSEGV, Segmentation fault.
0x0000555555e035e3 in std::__atomic_base<int>::operator++ (this=0x6c006c00650074) at /usr/include/c++/11/bits/atomic_base.h:385
385           { return __atomic_add_fetch(&_M_i, 1, int(memory_order_seq_cst)); }

10110111 · 2023-12-09T11:58:05Z

What's the stack trace?

alex-w · 2023-12-09T12:01:43Z

Next iteration in same steps:

LandscapeMgr: Clearing cache of 0 landscapes totalling about  0 MB.
StelGLWidget destroyed
[Thread 0x7fffdf298640 (LWP 138397) exited]

Thread 1 "stellarium" received signal SIGSEGV, Segmentation fault.
0x0000000000000000 in ?? ()

gzotti · 2023-12-09T12:30:55Z

Ah, now running debug mode build in debug run. Exited, and was pointed to an issue in a file atomic.

Debugger shows this function (with mark arrow)

  _NODISCARD _TVal load(const memory_order _Order) const noexcept { // load with given memory order
        const auto _Mem = _Atomic_address_as<int>(_Storage);
-->     auto _As_bytes  = __iso_volatile_load32(_Mem);
        _Load_barrier(_Order);
        return reinterpret_cast<_TVal&>(_As_bytes);
    }

I admit I have seriously no idea where this comes from. But such error in "atomic" was also the error when I gave up on the latest discussion about TelescopeControl.

Is the call display helpful:

Anything problematic in removing a StelTexture?

10110111 · 2023-12-09T13:48:39Z

I suppose we need to make a separate issue for tracking this crash. I've caught it once now in release mode run under Valgrind, with the same stack trace (in master). Apparently something is going wrong while deleting StelTextureMgr. From your stack trace it looks like StelTextureMgr::idMap was being deleted when the crash occurred.

gzotti · 2023-12-09T14:50:48Z

Maybe this is #3427 after all?

github-actions · 2023-12-10T14:03:37Z

Hello @gzotti!

Please check the fresh version (development snapshot) of Stellarium:
https://github.com/Stellarium/stellarium-data/releases/tag/weekly-snapshot

github-actions · 2023-12-23T11:53:39Z

Hello @gzotti!

Please check the latest stable version of Stellarium:
https://github.com/Stellarium/stellarium/releases/latest

Satellites: Restore shadow circle at arbitrary altitude

0412a9c

- also resets cutoff to 80km - also no commLink info for nonoperational satellites.

gzotti added the feature Entirely new feature label Dec 8, 2023

gzotti added this to the 23.4 milestone Dec 8, 2023

gzotti self-assigned this Dec 8, 2023

github-actions bot requested review from 10110111 and alex-w December 8, 2023 23:24

alex-w previously approved these changes Dec 9, 2023

View reviewed changes

Satellites: remove contradicting shadow enlargement

9ca1771

alex-w self-requested a review December 9, 2023 11:24

alex-w approved these changes Dec 9, 2023

View reviewed changes

gzotti marked this pull request as ready for review December 9, 2023 12:47

github-actions bot requested a review from alex-w December 9, 2023 12:47

Satellites: Avoid empty info subtitle

1af4db3

Satellites: SUG updates

0df1373

gzotti merged commit e4d9cc8 into master Dec 9, 2023
17 of 20 checks passed

gzotti deleted the fix/Sat-shadow-circle branch December 9, 2023 13:58

10110111 mentioned this pull request Dec 9, 2023

23.3: crash upon exit #3427

Closed

alex-w added the state: published The fix has been published for testing in weekly binary package label Dec 10, 2023

alex-w removed the state: published The fix has been published for testing in weekly binary package label Dec 23, 2023

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Satellites: Restore shadow circle at arbitrary altitude #3538

Satellites: Restore shadow circle at arbitrary altitude #3538

gzotti commented Dec 8, 2023 •

edited

Loading

alex-w commented Dec 9, 2023

gzotti commented Dec 9, 2023

alex-w commented Dec 9, 2023 •

edited

Loading

gzotti commented Dec 9, 2023

alex-w commented Dec 9, 2023

alex-w commented Dec 9, 2023

10110111 commented Dec 9, 2023

alex-w commented Dec 9, 2023

10110111 commented Dec 9, 2023

alex-w commented Dec 9, 2023

10110111 commented Dec 9, 2023

alex-w commented Dec 9, 2023

gzotti commented Dec 9, 2023 •

edited

Loading

10110111 commented Dec 9, 2023

gzotti commented Dec 9, 2023

github-actions bot commented Dec 10, 2023

github-actions bot commented Dec 23, 2023

Satellites: Restore shadow circle at arbitrary altitude #3538

Satellites: Restore shadow circle at arbitrary altitude #3538

Conversation

gzotti commented Dec 8, 2023 • edited Loading

Description

Screenshots (if appropriate):

Type of change

How Has This Been Tested?

Checklist:

alex-w commented Dec 9, 2023

gzotti commented Dec 9, 2023

alex-w commented Dec 9, 2023 • edited Loading

gzotti commented Dec 9, 2023

alex-w commented Dec 9, 2023

alex-w commented Dec 9, 2023

10110111 commented Dec 9, 2023

alex-w commented Dec 9, 2023

10110111 commented Dec 9, 2023

alex-w commented Dec 9, 2023

10110111 commented Dec 9, 2023

alex-w commented Dec 9, 2023

gzotti commented Dec 9, 2023 • edited Loading

10110111 commented Dec 9, 2023

gzotti commented Dec 9, 2023

github-actions bot commented Dec 10, 2023

github-actions bot commented Dec 23, 2023

gzotti commented Dec 8, 2023 •

edited

Loading

alex-w commented Dec 9, 2023 •

edited

Loading

gzotti commented Dec 9, 2023 •

edited

Loading