When trying to compile the attached file for stm8 or z80, sdcc from current trunk crashes. The file is manually simplified code originally created by csmith.
I see the issue for -mz80, -mstm8. -mhc08 --stack-auto. But not for -mpdk14, -mmcs51, -mhc08.
So apparently we get this problem when local variables are on the stack.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
SourceForge marks tickets as related when they are mentioned in each other's discussion. In this case, this happened since [bugs:#3483] was discovered using a code sample originally meant to reproduce this bug.
Original attachment avoids the reported crash on Z80, STM8, MCS51 stack-auto, PDK15 stack-auto and HC08 stack-auto. Other warnings remain. Failure phase not established as frontend-specific. https://godbolt.org/z/qqfrTq851
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
I see the issue for -mz80, -mstm8. -mhc08 --stack-auto. But not for -mpdk14, -mmcs51, -mhc08.
So apparently we get this problem when local variables are on the stack.
The main issue is fixed in [r13692], however there are apparently further issues that affect -mmcs51 --stack-auto and -mpdk15 --stack-auto.
Related
Commit: [r13692]
Last edit: Philipp Klaus Krause 2022-09-06
Testing with current SDCC [r12904], apparently this works now for -mpdk15 --stack-auto.
Bug #3483 is fixed. how does it relate to this one?
SourceForge marks tickets as related when they are mentioned in each other's discussion. In this case, this happened since [bugs:#3483] was discovered using a code sample originally meant to reproduce this bug.
Related
Bugs:
#3483Original attachment avoids the reported crash on Z80, STM8, MCS51 stack-auto, PDK15 stack-auto and HC08 stack-auto. Other warnings remain. Failure phase not established as frontend-specific.
https://godbolt.org/z/qqfrTq851
Updated regression test in [r17060]
Related
Commit: [r17060]