On amd64 and arm64 in order to walk the stack of an LNXDRV process we need to parse the DWARF metadata of the ELF image. Now we don't need this to handle SEH (ELF has no SEH handlers) but it would be nice to show a proper stack trace when an LNXDRV process crashes. Right now RtlVirtualUnwind will return error if it finds an ELF image instead of a PE image. What we could do is to call the DWARD stack walker when we detect that the instruction pointer belongs to an ELF module.
On amd64 and arm64 in order to walk the stack of an LNXDRV process we need to parse the DWARF metadata of the ELF image. Now we don't need this to handle SEH (ELF has no SEH handlers) but it would be nice to show a proper stack trace when an LNXDRV process crashes. Right now RtlVirtualUnwind will return error if it finds an ELF image instead of a PE image. What we could do is to call the DWARD stack walker when we detect that the instruction pointer belongs to an ELF module.