David Gibson <[email protected]> writes: > On Fri, Jun 08, 2018 at 12:54:36AM -0300, Philippe Mathieu-Daudé wrote: >> Hi David, >> >> On 06/08/2018 12:03 AM, David Gibson wrote: >> > On Thu, Jun 07, 2018 at 11:46:43AM -0300, Philippe Mathieu-Daudé wrote: >> >> Use abort() instead of error_setg(&error_abort), >> >> as suggested by the "qapi/error.h" documentation: >> >> >> >> Please don't error_setg(&error_fatal, ...), use error_report() and >> >> exit(), because that's more obvious. >> >> Likewise, don't error_setg(&error_abort, ...), use assert(). >> >> >> >> Use abort() instead of the suggested assert() because the assertion is >> >> already verified by the switch case. >> > >> > I think g_assert_not_reached() would be the right thing here. >> >> I try to follow Eric advice (who recalled Markus). >> As I understand: >> http://lists.gnu.org/archive/html/qemu-devel/2017-08/msg03605.html >> >> "glib-Testing [...] should not be used outside of tests/." > > Oh, ok, go with that then.
Most g_assert_FOO() are indeed tests-only, but g_assert_not_reached() is actually acceptable elsewhere, see commit 6e9389563e5, and its review thread https://lists.gnu.org/archive/html/qemu-devel/2017-04/msg05499.html Message-Id: <[email protected]> That said, I fail to see the value of this kind of lipstick, too. > Acked-by: David Gibson <[email protected]> > >> I can respin if you prefer. I can replace by g_assert_not_reached() on commit if you prefer.
