On 12/17/20 8:25 AM, Nathan Sidwell wrote:
On 12/17/20 5:07 AM, Jakub Jelinek wrote:
On Thu, Dec 17, 2020 at 11:01:40AM +0100, Andreas Schwab wrote:
/tmp/ccabCPZ7.s: Assembler messages:
/tmp/ccabCPZ7.s:28: Error: Wrong number of input operands
Yeah. If it is meant as an optimization barrier, shouldn't it be just
__asm__ volatile ("");
or
__asm__ volatile ("" : : : "memory");
Pushing to trunk:
This asm was a useful place for gdb to drop a breakpoint and make it
clear where you were when debugging. I took a punt that 'surely every
arch has a nop instruction'. Well, no, some apparently have nops with
operands (what, do nothing harder? :)
libcody/
* fatal.cc (HCF): Remove nop breakpoint lander.
--
Nathan Sidwell
diff --git i/libcody/fatal.cc w/libcody/fatal.cc
index b35094e6b19..c1bc8ab8a79 100644
--- i/libcody/fatal.cc
+++ w/libcody/fatal.cc
@@ -30,7 +30,8 @@ void (HCF) (char const *msg
#endif
) noexcept
{ // HCF - you goofed!
- __asm__ volatile ("nop"); // HCF - you goofed!
+ // A useful place for a breakpoint to land.
+ //__asm__ volatile ("nop"); // HCF - you goofed!
#if !NMS_CHECKING
constexpr Location loc (nullptr, 0);
