On Thu, 9 Sep 2004 11:39:41 -0700, Kent Sandvik <[EMAIL PROTECTED]> wrote:
> On Thu, 9 Sep 2004 04:39:54 +0200, Denis Oliver Kropp <[EMAIL PROTECTED]> wrote:
> Hmm. As it's hard to know what happens... Maybe it's OK to filter any
> low-end 'fake addresses' to avoid glitches.
And here's the temp patch I'm using just now. This is mostly to the
email records in case someone else has problems with bad info passed
to the signal_handler in future.
diff -ruN DirectFB-0.9.21/lib/direct/signals.c
DirectFB-0.9.21.mod/lib/direct/signals.c
--- DirectFB-0.9.21/lib/direct/signals.c 2004-08-04 09:58:18.000000000 -0700
+++ DirectFB-0.9.21.mod/lib/direct/signals.c 2004-09-09 11:50:43.000000000 -0700
@@ -330,7 +330,9 @@
fprintf( stderr, "(!) [%5d: %4lld.%03lld] --> Caught signal %d",
pid, millis/1000, millis%1000, num );
- if (info) {
+#define LOW_MEMORY_ADDRESS 0x100
+
+ if (info && info > LOW_MEMORY_ADDRESS) {
bool shown = false;
if (info->si_code > 0 && info->si_code < 0x80) {
Kent
--
Something, Anything!
