feedc0de/qt-creator
1
0
Fork 0
forked from qt-creator/qt-creator
Code Pull Requests Activity

be more lenient when matching the stops at _start

Browse Source 
archer reports stops at tbreaks properly, so checking for an empty stop
reason is no particularly good idea. instead, we simply skip *all* stops
at the entry point, assuming the user will not really set a breakpoint
there anyway.

Reviewed-by: hjk
This commit is contained in:
Oswald Buddenhagen
2009-11-11 11:34:16 +01:00
parent e4b1a25dd2
commit 5bd5fa63bf
1 changed files with 13 additions and 16 deletions
Download Patch File Download Diff File Expand all files Collapse all files

19
src/plugins/debugger/gdb/gdbengine.cpp
View File

@@ -1184,23 +1184,20 @@ void GdbEngine::handleStopResponse(const GdbMi &data)
if (!m_entryPoint.isEmpty()) {
GdbMi frameData = data.findChild("frame");
if (frameData.findChild("addr").data() == m_entryPoint) {
if (reason == "signal-received"
&& data.findChild("signal-name").data() == "SIGSTOP") {
// For some reason, attaching to a stopped process causes *two* stops
// There are two expected reasons for getting here:
// 1) For some reason, attaching to a stopped process causes *two* SIGSTOPs
// when trying to continue (kernel i386 2.6.24-23-ubuntu, gdb 6.8).
// Interestingly enough, on MacOSX no signal is delivered at all.
continueInferiorInternal();
return;
}
if (reason.isEmpty()) { // tbreak does that
// For programs without -pthread under gdb <= 6.8.
if (!inferiorPid())
// 2) The explicit tbreak at the entry point we set to query the PID.
// Gdb <= 6.8 reports a frame but no reason, 6.8.50+ reports everything.
// The case of the user really setting a breakpoint at _start is simply
// unsupported.
if (!inferiorPid()) // For programs without -pthread under gdb <= 6.8.
postCommand(_("info proc"), CB(handleInfoProc));
continueInferiorInternal();
return;
}
}
// We are past the initial stops. No need to waste time on further checks.
// We are past the initial stop(s). No need to waste time on further checks.
m_entryPoint.clear();
}
#endif