Re: [Lldb-commits] [lldb] r267478 - Fix StackFrame::GetVariables(...) function that was broken by 261858 when lambda functions were added to Block::AppendBlockVariables(). The Stackframe::GetVariables

2016-04-25 Thread Jim Ingham via lldb-commits
Debug mode doesn't mean no optimization, but -O0 SHOULD... OTOH, long experience has shown that it is very hard to convince compilers to be as dumb at -O0 as you would like them to be. That's why in all our test cases, we initialize variables, and generally do something dumb like printf the va

Re: [Lldb-commits] [lldb] r267478 - Fix StackFrame::GetVariables(...) function that was broken by 261858 when lambda functions were added to Block::AppendBlockVariables(). The Stackframe::GetVariables

2016-04-25 Thread Chaoren Lin via lldb-commits
Because it's compiled in debug mode? I wouldn't expect *any* optimization to be done in that case. On Mon, Apr 25, 2016 at 4:35 PM, Siva Chandra wrote: > Clang is probably right. Why bother to emit an unused file static scalar? > > On Mon, Apr 25, 2016 at 4:21 PM, Chaoren Lin wrote: > > Ah, I s

Re: [Lldb-commits] [lldb] r267478 - Fix StackFrame::GetVariables(...) function that was broken by 261858 when lambda functions were added to Block::AppendBlockVariables(). The Stackframe::GetVariables

2016-04-25 Thread Siva Chandra via lldb-commits
Clang is probably right. Why bother to emit an unused file static scalar? On Mon, Apr 25, 2016 at 4:21 PM, Chaoren Lin wrote: > Ah, I see. Is that a known clang bug? > > On Mon, Apr 25, 2016 at 4:07 PM, Siva Chandra > wrote: >> >> AFAICT, happens only with Clang; Using g_static_var, say as "retu

Re: [Lldb-commits] [lldb] r267478 - Fix StackFrame::GetVariables(...) function that was broken by 261858 when lambda functions were added to Block::AppendBlockVariables(). The Stackframe::GetVariables

2016-04-25 Thread Chaoren Lin via lldb-commits
Ah, I see. Is that a known clang bug? On Mon, Apr 25, 2016 at 4:07 PM, Siva Chandra wrote: > AFAICT, happens only with Clang; Using g_static_var, say as "return > g_static_var - 123;", fixes for me. > > On Mon, Apr 25, 2016 at 3:50 PM, Chaoren Lin via lldb-commits > wrote: > > Is g_global_var n

Re: [Lldb-commits] [lldb] r267478 - Fix StackFrame::GetVariables(...) function that was broken by 261858 when lambda functions were added to Block::AppendBlockVariables(). The Stackframe::GetVariables

2016-04-25 Thread Siva Chandra via lldb-commits
AFAICT, happens only with Clang; Using g_static_var, say as "return g_static_var - 123;", fixes for me. On Mon, Apr 25, 2016 at 3:50 PM, Chaoren Lin via lldb-commits wrote: > Is g_global_var necessarily static? > > On Linux, we're only seeing 2 static variables when your test expects 3. > >> sta

Re: [Lldb-commits] [lldb] r267478 - Fix StackFrame::GetVariables(...) function that was broken by 261858 when lambda functions were added to Block::AppendBlockVariables(). The Stackframe::GetVariables

2016-04-25 Thread Chaoren Lin via lldb-commits
Is g_global_var necessarily static? On Linux, we're only seeing 2 static variables when your test expects 3. > static_names = ['static_var', 'g_global_var', 'static_var'] I'm guessing g_global_var isn't treated as static. On Mon, Apr 25, 2016 at 2:54 PM, Greg Clayton via lldb-commits < lldb-co

[Lldb-commits] [lldb] r267478 - Fix StackFrame::GetVariables(...) function that was broken by 261858 when lambda functions were added to Block::AppendBlockVariables(). The Stackframe::GetVariables(...

2016-04-25 Thread Greg Clayton via lldb-commits
Author: gclayton Date: Mon Apr 25 16:54:10 2016 New Revision: 267478 URL: http://llvm.org/viewvc/llvm-project?rev=267478&view=rev Log: Fix StackFrame::GetVariables(...) function that was broken by 261858 when lambda functions were added to Block::AppendBlockVariables(). The Stackframe::GetVariab