Crash by entering consecutively in a measure quadruplets and other upper tuplets

• Jul 3, 2017 - 21:36
Reported version
S2 - Critical

2.1 version 871c8ce / Windows10

1) "Untitled" score
2) Select a measure
3) Hit four times Ctrl +4 (for quadruplet) or Ctrl + 5 (for quintuplet), and others (Ctrl + 6, etc.)

Result: crash

- Same results with a 3/4 time sig. But not with a 2/4 time sig.

- With Ctrl + 3 (default 4/4 time sig) and after pressing this command eight times, you get a warning "cannot create tuplet: note value is too short"
Hit "ok" -> Save / close / Reload
Result: you get a corruption.


With the 2.0.3, the behavior is slightly different.
You get warnings with quadruplets and others (after hitting the command 5 times)
But after "ok", and save/close/reload, you receive a crash with quadruplets, and corruption with quintuplets, sextuplets.

After checking in the recent past: I observe a change on March 8, 2017.

► With this nightly: 8ddd45d, I receive warnings with Ctrl + 4 (and upper) and Ctrl + 3. But after Ok, crash when save/close/restart

► With this nightly: 15a2ba4, I get crashs immediatly with Ctrl + 4 and Ctrl + 3

So, probably a side effect of this set of commits? :

▬ We observe the actual result (ie immediate crashs always with Ctrl + 4 and upper, and warning with Ctrl + 3, but corruption after reload) some days later, on March 21.

Ie, consecutively to this commit:…
To fix: #181886: Regression: Continuous tuplet creating on same place causes crash

- So, to sum up: Ctrl + 3 has been fixed (can we tell incompletely if considering corruption after restarting? )
For Ctrl + 4 and upper, the issue remains complete.

Status (old) patch (code needs review) fixed
Status fixed

Fixed in branch 2.2, commit efe9f23445

fix #268293: crash on creating tuplet with baseLen less than 1/128th
fix #229686: crash/corruption by consecutively entering upper tuplets