[Bug 287050] Change in PTRACE_CONTINUE causing valgrind/vgdb to no longer be able to interrupt debuggee

From: <bugzilla-noreply_at_freebsd.org>
Date: Thu, 29 May 2025 06:28:40 UTC
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=287050

--- Comment #12 from Paul Floyd <pjfloyd@wanadoo.fr> ---
(In reply to Konstantin Belousov from comment #11)

Things are looking better now.

The main issue now is that on aarch64 I get an assert. The vgdb invoker puts
0x8badf00d into x0 and then on the Valgrind side the VG_(invoke_gdbserver)
function contains an assert that its one argument has this value. Instead it
has a value of 4 so that assert fires. I'm not sure what is happening -
ptrace(PT_SETREGS, pid, (caddr_t)regs, 0) isn't doing its job or x0 is getting
clobbered somehow between the PT_SETREGS ans the PT_CONTINUE.

-- 
You are receiving this mail because:
You are the assignee for the bug.