You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
RuntimeError: memory access out of bounds
at runtime.morestack (wasm://wasm/0370d90e:wasm-function[1436]:0x134e6f)
at runtime.morestack_noctxt (wasm://wasm/0370d90e:wasm-function[1437]:0x134eab)
at runtime.newstack (wasm://wasm/0370d90e:wasm-function[977]:0xe4596)
at runtime.morestack (wasm://wasm/0370d90e:wasm-function[1436]:0x134e8d)
at runtime.morestack_noctxt (wasm://wasm/0370d90e:wasm-function[1437]:0x134eab)
at runtime.newstack (wasm://wasm/0370d90e:wasm-function[977]:0xe4596)
at runtime.morestack (wasm://wasm/0370d90e:wasm-function[1436]:0x134e8d)
at runtime.morestack_noctxt (wasm://wasm/0370d90e:wasm-function[1437]:0x134eab)
at runtime.newstack (wasm://wasm/0370d90e:wasm-function[977]:0xe4596)
at runtime.morestack (wasm://wasm/0370d90e:wasm-function[1436]:0x134e8d)
With the fix applied:
RuntimeError: memory access out of bounds
at runtime.morestack (wasm://wasm/0370d90e:wasm-function[1436]:0x134e6f)
at runtime.morestack_noctxt (wasm://wasm/0370d90e:wasm-function[1437]:0x134eab)
at runtime.write1 (wasm://wasm/0370d90e:wasm-function[662]:0x9c78f)
at runtime.write (wasm://wasm/0370d90e:wasm-function[1056]:0xfbefc)
at runtime.badmorestackg0 (wasm://wasm/0370d90e:wasm-function[766]:0xad590)
at runtime.morestack (wasm://wasm/0370d90e:wasm-function[1436]:0x134df1)
at runtime.morestack_noctxt (wasm://wasm/0370d90e:wasm-function[1437]:0x134eab)
at runtime.write1 (wasm://wasm/0370d90e:wasm-function[662]:0x9c78f)
at runtime.write (wasm://wasm/0370d90e:wasm-function[1056]:0xfbefc)
at runtime.badmorestackg0 (wasm://wasm/0370d90e:wasm-function[766]:0xad590)
While investigating crash stack on wasm, i found that badmorestackg0 is behind a condition that is never true.
go/src/runtime/asm_wasm.s
Lines 217 to 230 in d72f454
As you can see above, it is comparing g == R1, R1 being g.m. We should be comparing it to g0, which is stored in R2.
It is also present in 1.20 and 1.21, but I'm not sure if we need to backport this given that it is not a significant issue.
The text was updated successfully, but these errors were encountered: