gcc-15 
 07: 
  On 2025-11-02 Andreas Metzler  wrote: 
 > [...] 
 > > Also setting either no-schedule-insns2 or no-strict-aliasing for this 
 > > function (instead of O1) for a -O2 build lets the arm64 build succeed. 
 > 
 > or alternatively 
 > -O1  -fschedule-insns2 -fstrict-aliasing 
 > breaks the build. 
  
 The segfault also does not occur with -O3, and the important difference 
 seems to be that -O3 enables -fschedule-insns (without trailing '2'). 
  
 Building with -O2 -fschedule-insns fixes pamtilt. 
  
 --- SoupGate-Win32 v1.05 
  * Origin: you cannot sedate... all the things you hate (1:229/2)