[PATCH v4] Fix range end handling of inlined subroutines - restart

[Bernd Edlinger]

Hi,

this is a new improved version of my patch.

It addresses two issues that came up when playing and debugging
the rather big and complex gcc application.

First I found cases where a completely empty inlined subroutine
was encountered, the step command can step into the inline,
but the call stack does not reflect that.  Fixed that by
making an exception in dwarf2_get_pc_bounds for empty inlined
routines.

Then if a function is called immediatly after the inline,
the call stack shows the last line of the inline, but we
have a better line, a !is-stmt one following the inline
function.  Fixed that by making find_pc_sect_line prefer a
!is_stmt line over a weak one if an in-between PC is requested.

There is a new test case for those two issues.

Finally I found that a "b inline-function" does not
always find the entry_pc, and instead the breakpoint is
at the first non-zero block range of the inline block.

Fixed that by moving the block that starts at entry_pc
to the first position.


Tested successfully on x86_64-pc-linux-gnu.
Is it OK for trunk?


Thanks
Bernd.
 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 0001-Fix-range-end-handling-of-inlined-subroutines.patch
Type: text/x-patch
Size: 35817 bytes
Desc: not available
URL: <https://sourceware.org/pipermail/gdb-patches/attachments/20201129/fec0b5e9/attachment-0001.bin>