There's a misunderstanding. https://github.com/pure-data/pure-data/commit/95b614656cd62214d6779014e5173d8af697814b is the one that switched to stack allocation.So if it was commit 95b614656cd62214d6779014e5173d8af697814b was Aug 2021,
The commit that (very likely) causes your crash is https://github.com/pure-data/pure-data/commit/150858a65e7f580a32b39746338c0fb163faa80b resp. https://github.com/pure-data/pure-data/commit/3539e1d8d4d64149c2507527cfc44428246d34b4.
So can you please try the latest develop branch and report if it fixes your issue?
Christof
Thanks for the find I now confirm it's Version 0.55 where I found it, so I changed Subject. So if it was commit 95b614656cd62214d6779014e5173d8af697814b was Aug 2021, and I had it only after Debian bookworm-backports upgrade. Maybe there is something else which changed. Interims solution: downgrade to stable and use 0.53.1. mfg winfried PS.: My gut feeling tells me, it would be better one allocation, since within an installation a lot of RAM can be grabbed by others making it more erratic behaviour, but I'm not that deeply familiar with the "threads" matte . Am Dienstag, 8. Oktober 2024, 23:28:44 CEST schrieb IOhannes m zmölnig:On 08/10/2024 22:04, Christof Ressi wrote:@IOhannes: why don't we just allocate the buffer on the heap *once* before we start the stream?no idea. but 95b614656cd62214d6779014e5173d8af697814b wasn't me :-)gfmdsaf IOhannes