https://bugs.kde.org/show_bug.cgi?id=524048
Harald Sitter <[email protected]> changed: What |Removed |Added ---------------------------------------------------------------------------- Latest Commit|https://invent.kde.org/plas |https://invent.kde.org/plas |ma/drkonqi/-/commit/8e63a84 |ma/drkonqi/-/commit/c8d6939 |ddab6884564d9e56ad86e1e508e |5bea93e913d45ff438564ebd548 |87a4a0 |4221db --- Comment #5 from Harald Sitter <[email protected]> --- Git commit c8d69395bea93e913d45ff438564ebd5484221db by Harald Sitter. Committed on 28/08/2026 at 10:17. Pushed by sitter into branch 'Plasma/6.7'. coredump: don't process crashes of our own binaries A crash of the launcher is picked up by systemd-coredump like any other one, gets processed, and activates another launcher. When that one crashes as well the pipeline starts feeding itself. On the reporter's machine it ran at about 2.5 iterations per second and piled up ~2 GB of compressed cores in four and a half minutes, all from three Spectacle crashes. Drop dumps whose executable is one of ours before anything gets forwarded. Our own crashes are still in the journal and in systemd-coredump's storage, they just don't go around the loop once more. (cherry picked from commit 8e63a84ddab6884564d9e56ad86e1e508e87a4a0) Co-authored-by: Alexander Berg <[email protected]> M +29 -0 src/coredump/processor/main.cpp https://invent.kde.org/plasma/drkonqi/-/commit/c8d69395bea93e913d45ff438564ebd5484221db -- You are receiving this mail because: You are watching all bug changes.
