Mirror of the gdb-patches mailing list
 help / color / mirror / Atom feed
From: Andrew Burgess <andrew.burgess@embecosm.com>
To: Bernd Edlinger <bernd.edlinger@hotmail.de>
Cc: gdb-patches@sourceware.org, Tom Tromey <tom@tromey.com>
Subject: Re: [PATCH 2/2] gdb: Preserve is-stmt lines when switch between files
Date: Sat, 4 Apr 2020 23:23:26 +0100	[thread overview]
Message-ID: <20200404222326.GB3917@embecosm.com> (raw)
In-Reply-To: <PR2PR03MB5179B22E4BBD59D964CCC77CE4C40@PR2PR03MB5179.eurprd03.prod.outlook.com>

* Bernd Edlinger <bernd.edlinger@hotmail.de> [2020-04-04 20:07:52 +0200]:

> Sorry, Andrew,
> 
> please hold this one.
> 
> This will definitely break all the step over inline
> stuff that was already working.

By all I assume you only mean gdb.cp/step-and-next-inline.exp, which I
specifically talk about in the commit message.

The DWARF produced in that test is horribly broken, I was tempted to
drop the whole test, but a few parts do still work, and hopefully if
GCC gets fixed we'll be back in business.

As I mentioned in the GCC bug report, if I hack GDB to "fix" the DWARF
produced by GCC, then that test passes fine with no regressions with
this patch as is.

All other tests still pass with this patch - admittedly there's not
much testing the inline behaviour, but there are a few.

If you know of a case with valid DWARF that is broken by this test
then I'm all ears.  With the setup I used for the tests in this patch
it's actually pretty easy to synthesis test of inline functions, so
all you really need to do is describe a test you think will break and
I can knock up a test case for it reasonably quickly.

> 
> Can we please go to the drawing board before we continue
> from here?

Sure.  I posted my thoughts just now.

Thanks,
Andrew

> 
> 
> Thanks
> Bernd.
> 
> On 4/4/20 12:21 AM, Andrew Burgess wrote:
> > After the is-stmt support commit:
> > 
> >   commit 8c95582da858ac981f689a6f599acacb8c5c490f
> >   Date:   Mon Dec 30 21:04:51 2019 +0000
> > 
> >       gdb: Add support for tracking the DWARF line table is-stmt field
> > 
> > A regression was observed where a breakpoint could no longer be placed
> > in some cases.
> > 
> > Consider a line table like this:
> > 
> >   File 1: test.c
> >   File 2: test.h
> > 
> >   | Addr | File | Line | Stmt |
> >   |------|------|------|------|
> >   | 1    | 1    | 16   | Y    |
> >   | 2    | 1    | 17   | Y    |
> >   | 3    | 2    | 21   | Y    |
> >   | 4    | 2    | 22   | Y    |
> >   | 4    | 1    | 18   | N    |
> >   | 5    | 2    | 23   | N    |
> >   | 6    | 1    | 24   | Y    |
> >   | 7    | 1    | END  | Y    |
> >   |------|------|------|------|
> > 
> > Before the is-stmt patch GDB would ignore any non-stmt lines, so GDB
> > built two line table structures:
> > 
> >   File 1                 File 2
> >   ------                 ------
> > 
> >   | Addr | Line |        | Addr | Line |
> >   |------|------|        |------|------|
> >   | 1    | 16   |        | 3    | 21   |
> >   | 2    | 17   |        | 4    | 22   |
> >   | 3    | END  |        | 6    | END  |
> >   | 6    | 24   |        |------|------|
> >   | 7    | END  |
> >   |------|------|
> > 
> > After the is-stmt patch GDB now records non-stmt lines, so the
> > generated line table structures look like this:
> > 
> >   File 1                   File 2
> >   ------                   ------
> > 
> >   | Addr | Line | Stmt |  | Addr | Line | Stmt |
> >   |------|------|------|  |------|------|------|
> >   | 1    | 16   | Y    |  | 3    | 21   | Y    |
> >   | 2    | 17   | Y    |  | 4    | 22   | Y    |
> >   | 3    | END  | Y    |  | 4    | END  | Y    |
> >   | 4    | 18   | N    |  | 5    | 23   | N    |
> >   | 5    | END  | Y    |  | 6    | END  | Y    |
> >   | 6    | 24   | Y    |  |------|------|------|
> >   | 7    | END  | Y    |
> >   |------|------|------|
> > 
> > The problem is that in 'File 2', end END marker at address 4 causes
> > the previous line table entry to be discarded, so we actually end up
> > with this:
> > 
> >   File 2
> >   ------
> > 
> >   | Addr | Line | Stmt |
> >   |------|------|------|
> >   | 3    | 21   | Y    |
> >   | 4    | END  | Y    |
> >   | 5    | 23   | N    |
> >   | 6    | END  | Y    |
> >   |------|------|------|
> > 
> > When a user tries to place a breakpoint in file 2 at line 22, this is
> > no longer possible.
> > 
> > The solution I propose here is that we ignore line table entries that
> > would trigger a change of file if:
> > 
> >   1. The new line being added is at the same address as the previous
> >   line, and
> > 
> >   2. We have previously seen an is-stmt line at the current address.
> > 
> > The result of this is that GDB switches file, and knows that some line
> > entry (or entries) are going to be discarded, prefer to keep is-stmt
> > lines and discard non-stmt lines.
> > 
> > After this commit the lines tables are now:
> > 
> >   File 1                   File 2
> >   ------                   ------
> > 
> >   | Addr | Line | Stmt |  | Addr | Line | Stmt |
> >   |------|------|------|  |------|------|------|
> >   | 1    | 16   | Y    |  | 3    | 21   | Y    |
> >   | 2    | 17   | Y    |  | 4    | 22   | Y    |
> >   | 3    | END  | Y    |  | 5    | 23   | N    |
> >   | 5    | END  | Y    |  | 6    | END  | Y    |
> >   | 6    | 24   | Y    |  |------|------|------|
> >   | 7    | END  | Y    |
> >   |------|------|------|
> > 
> > We've lost the non-stmt entry for file 1, line 18, but retained the
> > is-stmt entry for file 2, line 22.  The user can now place a
> > breakpoint at that location.
> > 
> > One problem that came from this commit was the test
> > gdb.cp/step-and-next-inline.exp, which broke in several places.  After
> > looking at this test again I think that in some cases this test was
> > only ever passing by pure luck.  The debug GCC is producing for this
> > test is pretty broken.  I raised this GCC bug:
> > 
> >   https://gcc.gnu.org/bugzilla/show_bug.cgi?id=94474
> > 
> > for this and disabled one entire half of the test.  There are still
> > some cases in here that do pass, and if/when GCC is fixed it would be
> > great to enable this test again.
> > 
> > gdb/ChangeLog:
> > 
> > 	* dwarf2/read.c (class lnp_state_machine) <m_last_address>: New
> > 	member variable.
> > 	<m_stmt_at_address>: New member variable.
> > 	(lnp_state_machine::record_line): Don't record some lines, update
> > 	tracking of is_stmt at the same address.
> > 	(lnp_state_machine::lnp_state_machine): Initialise new member
> > 	variables.
> > 
> > gdb/testsuite/ChangeLog:
> > 
> > 	* gdb.cp/step-and-next-inline.exp (do_test): Skip all tests in the
> > 	use_header case.
> > 	* gdb.dwarf2/dw2-inline-header-1.exp: New file.
> > 	* gdb.dwarf2/dw2-inline-header-2.exp: New file.
> > 	* gdb.dwarf2/dw2-inline-header-3.exp: New file.
> > 	* gdb.dwarf2/dw2-inline-header-lbls.c: New file.
> > 	* gdb.dwarf2/dw2-inline-header.c: New file.
> > 	* gdb.dwarf2/dw2-inline-header.h: New file.
> > ---
> >  gdb/ChangeLog                                     |  10 ++
> >  gdb/dwarf2/read.c                                 |  47 +++++-
> >  gdb/testsuite/ChangeLog                           |  11 ++
> >  gdb/testsuite/gdb.cp/step-and-next-inline.exp     |   7 +
> >  gdb/testsuite/gdb.dwarf2/dw2-inline-header-1.exp  | 156 ++++++++++++++++++
> >  gdb/testsuite/gdb.dwarf2/dw2-inline-header-2.exp  | 179 ++++++++++++++++++++
> >  gdb/testsuite/gdb.dwarf2/dw2-inline-header-3.exp  | 192 ++++++++++++++++++++++
> >  gdb/testsuite/gdb.dwarf2/dw2-inline-header-lbls.c |  46 ++++++
> >  gdb/testsuite/gdb.dwarf2/dw2-inline-header.c      |  24 +++
> >  gdb/testsuite/gdb.dwarf2/dw2-inline-header.h      |  24 +++
> >  10 files changed, 693 insertions(+), 3 deletions(-)
> >  create mode 100644 gdb/testsuite/gdb.dwarf2/dw2-inline-header-1.exp
> >  create mode 100644 gdb/testsuite/gdb.dwarf2/dw2-inline-header-2.exp
> >  create mode 100644 gdb/testsuite/gdb.dwarf2/dw2-inline-header-3.exp
> >  create mode 100644 gdb/testsuite/gdb.dwarf2/dw2-inline-header-lbls.c
> >  create mode 100644 gdb/testsuite/gdb.dwarf2/dw2-inline-header.c
> >  create mode 100644 gdb/testsuite/gdb.dwarf2/dw2-inline-header.h
> > 
> > diff --git a/gdb/dwarf2/read.c b/gdb/dwarf2/read.c
> > index f94c66b4f1b..261c8455424 100644
> > --- a/gdb/dwarf2/read.c
> > +++ b/gdb/dwarf2/read.c
> > @@ -19327,6 +19327,15 @@ class lnp_state_machine
> >    /* The last file a line number was recorded for.  */
> >    struct subfile *m_last_subfile = NULL;
> >  
> > +  /* The address of the last line entry.  */
> > +  CORE_ADDR m_last_address;
> > +
> > +  /* Set to true when a previous line at the same address (using
> > +     m_last_address) had m_is_stmt true.  This is reset to false when a
> > +     line entry at a new address (m_address different to m_last_address) is
> > +     processed.  */
> > +  bool m_stmt_at_address = false;
> > +
> >    /* When true, record the lines we decode.  */
> >    bool m_currently_recording_lines = false;
> >  
> > @@ -19520,14 +19529,34 @@ lnp_state_machine::record_line (bool end_sequence)
> >        fe->included_p = 1;
> >        if (m_record_lines_p)
> >  	{
> > -	  if (m_last_subfile != m_cu->get_builder ()->get_current_subfile ()
> > -	      || end_sequence)
> > +	  /* When we switch files we insert an end maker in the first file,
> > +	     switch to the second file and add a new line entry.  The
> > +	     problem is that the end marker inserted in the first file will
> > +	     discard any previous line entries at the same address.  If the
> > +	     line entries in the first file are marked as is-stmt, while
> > +	     the new line in the second file is non-stmt, then this means
> > +	     the end marker will discard is-stmt lines so we can have a
> > +	     non-stmt line.  This means that there are less addresses at
> > +	     which the user can insert a breakpoint.
> > +
> > +	     To improve this we track the last address in m_last_address,
> > +	     and whether we have seen an is-stmt at this address.  Then
> > +	     when switching files, if we have seen a stmt at the current
> > +	     address, and we are switching to create a non-stmt line, then
> > +	     discard the new line.  */
> > +	  bool file_changed
> > +	    = m_last_subfile != m_cu->get_builder ()->get_current_subfile ();
> > +	  bool ignore_this_line
> > +	    = (file_changed && !end_sequence && m_last_address == m_address
> > +	       && !m_is_stmt && m_stmt_at_address);
> > +
> > +	  if ((file_changed && !ignore_this_line) || end_sequence)
> >  	    {
> >  	      dwarf_finish_line (m_gdbarch, m_last_subfile, m_address,
> >  				 m_currently_recording_lines ? m_cu : nullptr);
> >  	    }
> >  
> > -	  if (!end_sequence)
> > +	  if (!end_sequence && !ignore_this_line)
> >  	    {
> >  	      bool is_stmt = producer_is_codewarrior (m_cu) || m_is_stmt;
> >  
> > @@ -19546,6 +19575,15 @@ lnp_state_machine::record_line (bool end_sequence)
> >  	    }
> >  	}
> >      }
> > +
> > +  /* Track whether we have seen any m_is_stmt true at m_address in case we
> > +     have multiple line table entries all at m_address.  */
> > +  if (m_last_address != m_address)
> > +    {
> > +      m_stmt_at_address = false;
> > +      m_last_address = m_address;
> > +    }
> > +  m_stmt_at_address |= m_is_stmt;
> >  }
> >  
> >  lnp_state_machine::lnp_state_machine (struct dwarf2_cu *cu, gdbarch *arch,
> > @@ -19565,6 +19603,9 @@ lnp_state_machine::lnp_state_machine (struct dwarf2_cu *cu, gdbarch *arch,
> >    m_address = gdbarch_adjust_dwarf2_line (arch, 0, 0);
> >    m_is_stmt = lh->default_is_stmt;
> >    m_discriminator = 0;
> > +
> > +  m_last_address = m_address;
> > +  m_stmt_at_addr = false;
> >  }
> >  
> >  void
> > diff --git a/gdb/testsuite/gdb.cp/step-and-next-inline.exp b/gdb/testsuite/gdb.cp/step-and-next-inline.exp
> > index 3733fa75570..a95e21194f9 100644
> > --- a/gdb/testsuite/gdb.cp/step-and-next-inline.exp
> > +++ b/gdb/testsuite/gdb.cp/step-and-next-inline.exp
> > @@ -24,6 +24,13 @@ if { ![supports_statement_frontiers] } {
> >  proc do_test { use_header } {
> >      global srcfile testfile
> >  
> > +    if { $use_header } {
> > +	# This test will not pass due to poor debug information
> > +	# generated by GCC (at least upto 10.x).  See
> > +	# https://gcc.gnu.org/bugzilla/show_bug.cgi?id=94474
> > +	return
> > +    }
> > +
> >      set options {c++ debug nowarnings optimize=-O2\ -gstatement-frontiers}
> >      if { $use_header } {
> >  	lappend options additional_flags=-DUSE_NEXT_INLINE_H
> > diff --git a/gdb/testsuite/gdb.dwarf2/dw2-inline-header-1.exp b/gdb/testsuite/gdb.dwarf2/dw2-inline-header-1.exp
> > new file mode 100644
> > index 00000000000..6a1e990002c
> > --- /dev/null
> > +++ b/gdb/testsuite/gdb.dwarf2/dw2-inline-header-1.exp
> > @@ -0,0 +1,156 @@
> > +# Copyright 2020 Free Software Foundation, Inc.
> > +
> > +# This program is free software; you can redistribute it and/or modify
> > +# it under the terms of the GNU General Public License as published by
> > +# the Free Software Foundation; either version 3 of the License, or
> > +# (at your option) any later version.
> > +#
> > +# This program is distributed in the hope that it will be useful,
> > +# but WITHOUT ANY WARRANTY; without even the implied warranty of
> > +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
> > +# GNU General Public License for more details.
> > +#
> > +# You should have received a copy of the GNU General Public License
> > +# along with this program.  If not, see <http://www.gnu.org/licenses/>.
> > +
> > +# Setup a line table where:
> > +#
> > +# | Addr | File | Line | Stmt |
> > +# |------|------|------|------|
> > +# | 1    | 1    | 16   | Y    |
> > +# | 2    | 1    | 17   | Y    |
> > +# | 3    | 2    | 21   | Y    |
> > +# | 4    | 2    | 22   | Y    |
> > +# | 4    | 1    | 18   | N    |
> > +# | 5    | 2    | 23   | N    |
> > +# | 6    | 1    | 24   | Y    |
> > +# | 7    | 1    | END  | Y    |
> > +# |------|------|------|------|
> > +#
> > +# Places a brekpoint at file 2, line 22.  Previously GDB would discrad
> > +# the line table entry for this line due to switching files for the
> > +# file 1, line 18 non-statement line.  After patching however, GDB now
> > +# discards the file 1, line 18 entry instead, and the breakpoint at
> > +# line 22 should succeed.
> > +
> > +load_lib dwarf.exp
> > +
> > +# This test can only be run on targets which support DWARF-2 and use gas.
> > +if {![dwarf2_support]} {
> > +    return 0
> > +}
> > +
> > +# The .c files use __attribute__.
> > +if [get_compiler_info] {
> > +    return -1
> > +}
> > +if !$gcc_compiled {
> > +    return 0
> > +}
> > +
> > +standard_testfile dw2-inline-header-lbls.c dw2-inline-header.S \
> > +    dw2-inline-header.c dw2-inline-header.h
> > +
> > +set asm_file [standard_output_file $srcfile2]
> > +Dwarf::assemble $asm_file {
> > +    global srcdir subdir srcfile srcfile3 srcfile4
> > +    declare_labels lines_label callee_subprog_label
> > +
> > +    get_func_info main
> > +
> > +    cu {} {
> > +	compile_unit {
> > +	    {producer "gcc" }
> > +	    {language @DW_LANG_C}
> > +	    {name ${srcfile3}}
> > +	    {low_pc 0 addr}
> > +	    {stmt_list ${lines_label} DW_FORM_sec_offset}
> > +	} {
> > +	    callee_subprog_label: subprogram {
> > +		{external 1 flag}
> > +		{name callee}
> > +		{inline 3 data1}
> > +	    }
> > +	    subprogram {
> > +		{external 1 flag}
> > +		{name main}
> > +		{low_pc $main_start addr}
> > +		{high_pc "$main_start + $main_len" addr}
> > +	    } {
> > +		inlined_subroutine {
> > +		    {abstract_origin %$callee_subprog_label}
> > +		    {low_pc line_label_1 addr}
> > +		    {high_pc line_label_7 addr}
> > +		    {call_file 1 data1}
> > +		    {call_line 18 data1}
> > +		}
> > +	    }
> > +	}
> > +    }
> > +
> > +    lines {version 2 default_is_stmt 1} lines_label {
> > +	include_dir "${srcdir}/${subdir}"
> > +	file_name "$srcfile3" 1
> > +	file_name "$srcfile4" 1
> > +
> > +	program {
> > +	    {DW_LNE_set_address line_label_1}
> > +	    {DW_LNS_advance_line 15}
> > +	    {DW_LNS_copy}
> > +
> > +	    {DW_LNE_set_address line_label_2}
> > +	    {DW_LNS_advance_line 1}
> > +	    {DW_LNS_copy}
> > +
> > +	    {DW_LNS_set_file 2}
> > +	    {DW_LNE_set_address line_label_3}
> > +	    {DW_LNS_advance_line 4}
> > +	    {DW_LNS_copy}
> > +
> > +	    {DW_LNE_set_address line_label_4}
> > +	    {DW_LNS_advance_line 1}
> > +	    {DW_LNS_copy}
> > +
> > +	    {DW_LNS_advance_line -4}
> > +	    {DW_LNS_set_file 1}
> > +	    {DW_LNS_negate_stmt}
> > +	    {DW_LNS_copy}
> > +
> > +	    {DW_LNS_set_file 2}
> > +	    {DW_LNE_set_address line_label_5}
> > +	    {DW_LNS_advance_line 5}
> > +	    {DW_LNS_copy}
> > +
> > +	    {DW_LNS_negate_stmt}
> > +	    {DW_LNS_set_file 1}
> > +	    {DW_LNE_set_address line_label_6}
> > +	    {DW_LNS_advance_line 1}
> > +	    {DW_LNS_copy}
> > +
> > +	    {DW_LNE_set_address line_label_7}
> > +	    {DW_LNE_end_sequence}
> > +	}
> > +    }
> > +}
> > +
> > +if { [prepare_for_testing "failed to prepare" ${testfile} \
> > +	  [list $srcfile $asm_file] {nodebug optimize=-O1}] } {
> > +    return -1
> > +}
> > +
> > +if ![runto_main] {
> > +    return -1
> > +}
> > +
> > +# Delete all breakpoints so that the output of "info breakpoints"
> > +# below will only contain a single breakpoint.
> > +delete_breakpoints
> > +
> > +# Place a breakpoint within the function in the header file.
> > +gdb_breakpoint "${srcfile4}:22"
> > +
> > +# Check that the breakpoint was placed where we expected.  It should
> > +# appear at the requested line.  When the bug in GDB was present the
> > +# breakpoint would be placed on one of the following lines instead.
> > +gdb_test "info breakpoints" \
> > +    ".* in callee at \[^\r\n\]+${srcfile4}:22\\y.*"
> > diff --git a/gdb/testsuite/gdb.dwarf2/dw2-inline-header-2.exp b/gdb/testsuite/gdb.dwarf2/dw2-inline-header-2.exp
> > new file mode 100644
> > index 00000000000..46499919a8b
> > --- /dev/null
> > +++ b/gdb/testsuite/gdb.dwarf2/dw2-inline-header-2.exp
> > @@ -0,0 +1,179 @@
> > +# Copyright 2020 Free Software Foundation, Inc.
> > +
> > +# This program is free software; you can redistribute it and/or modify
> > +# it under the terms of the GNU General Public License as published by
> > +# the Free Software Foundation; either version 3 of the License, or
> > +# (at your option) any later version.
> > +#
> > +# This program is distributed in the hope that it will be useful,
> > +# but WITHOUT ANY WARRANTY; without even the implied warranty of
> > +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
> > +# GNU General Public License for more details.
> > +#
> > +# You should have received a copy of the GNU General Public License
> > +# along with this program.  If not, see <http://www.gnu.org/licenses/>.
> > +
> > +# Setup a line table where:
> > +#
> > +# | Addr | File | Line | Stmt |
> > +# |------|------|------|------|
> > +# | 1    | 1    | 16   | Y    |
> > +# | 2    | 1    | 17   | Y    |
> > +# | 3    | 2    | 21   | Y    |
> > +# | 4    | 2    | 22   | Y    |
> > +# | 4    | 1    | 18   | N    |
> > +# | 5    | 1    | 19   | Y    |
> > +# | 6    | 1    | 20   | Y    |
> > +# | 7    | 1    | END  | Y    |
> > +# |------|------|------|------|
> > +#
> > +#
> > +# Place the first brekpoint at file 2, line 22 and a second breakpoint
> > +# at file 1, line 19.  A third breakpoint is placed at file 1, line
> > +# 18, but as this line table entry will have been discarded[1] the
> > +# third breakpoint will actually be placed at the same location as the
> > +# second breakpoint.
> > +#
> > +# [1] The entry for file 1, line 18 is discarded because it is at the
> > +# same address as the previous entry, but the previous entry is-stmt,
> > +# while line 18 is a non-stmt.
> > +
> > +load_lib dwarf.exp
> > +
> > +# This test can only be run on targets which support DWARF-2 and use gas.
> > +if {![dwarf2_support]} {
> > +    return 0
> > +}
> > +
> > +# The .c files use __attribute__.
> > +if [get_compiler_info] {
> > +    return -1
> > +}
> > +if !$gcc_compiled {
> > +    return 0
> > +}
> > +
> > +standard_testfile dw2-inline-header-lbls.c dw2-inline-header.S \
> > +    dw2-inline-header.c dw2-inline-header.h
> > +
> > +set asm_file [standard_output_file $srcfile2]
> > +Dwarf::assemble $asm_file {
> > +    global srcdir subdir srcfile srcfile3 srcfile4
> > +    declare_labels lines_label callee_subprog_label
> > +
> > +    get_func_info main
> > +
> > +    cu {} {
> > +	compile_unit {
> > +	    {producer "gcc" }
> > +	    {language @DW_LANG_C}
> > +	    {name ${srcfile3}}
> > +	    {low_pc 0 addr}
> > +	    {stmt_list ${lines_label} DW_FORM_sec_offset}
> > +	} {
> > +	    callee_subprog_label: subprogram {
> > +		{external 1 flag}
> > +		{name callee}
> > +		{inline 3 data1}
> > +	    }
> > +	    subprogram {
> > +		{external 1 flag}
> > +		{name main}
> > +		{low_pc $main_start addr}
> > +		{high_pc "$main_start + $main_len" addr}
> > +	    } {
> > +		inlined_subroutine {
> > +		    {abstract_origin %$callee_subprog_label}
> > +		    {low_pc line_label_1 addr}
> > +		    {high_pc line_label_7 addr}
> > +		    {call_file 1 data1}
> > +		    {call_line 18 data1}
> > +		}
> > +	    }
> > +	}
> > +    }
> > +
> > +    lines {version 2 default_is_stmt 1} lines_label {
> > +	include_dir "${srcdir}/${subdir}"
> > +	file_name "$srcfile3" 1
> > +	file_name "$srcfile4" 1
> > +
> > +	program {
> > +	    {DW_LNE_set_address line_label_1}
> > +	    {DW_LNS_advance_line 15}
> > +	    {DW_LNS_copy}
> > +
> > +	    {DW_LNE_set_address line_label_2}
> > +	    {DW_LNS_advance_line 1}
> > +	    {DW_LNS_copy}
> > +
> > +	    {DW_LNS_set_file 2}
> > +	    {DW_LNE_set_address line_label_3}
> > +	    {DW_LNS_advance_line 4}
> > +	    {DW_LNS_copy}
> > +
> > +	    {DW_LNE_set_address line_label_4}
> > +	    {DW_LNS_advance_line 1}
> > +	    {DW_LNS_copy}
> > +
> > +	    {DW_LNS_advance_line -4}
> > +	    {DW_LNS_set_file 1}
> > +	    {DW_LNS_negate_stmt}
> > +	    {DW_LNS_copy}
> > +
> > +	    {DW_LNE_set_address line_label_5}
> > +	    {DW_LNS_advance_line 1}
> > +	    {DW_LNS_negate_stmt}
> > +	    {DW_LNS_copy}
> > +
> > +	    {DW_LNE_set_address line_label_6}
> > +	    {DW_LNS_advance_line 1}
> > +	    {DW_LNS_copy}
> > +
> > +	    {DW_LNE_set_address line_label_7}
> > +	    {DW_LNE_end_sequence}
> > +	}
> > +    }
> > +}
> > +
> > +if { [prepare_for_testing "failed to prepare" ${testfile} \
> > +	  [list $srcfile $asm_file] {nodebug optimize=-O1}] } {
> > +    return -1
> > +}
> > +
> > +if ![runto_main] {
> > +    return -1
> > +}
> > +
> > +# Delete all breakpoints so that the output of "info breakpoints"
> > +# below will only contain a single breakpoint.
> > +delete_breakpoints
> > +
> > +# Place a breakpoint within the function in the header file.
> > +gdb_breakpoint "${srcfile4}:22"
> > +
> > +# Check that the breakpoint was placed where we expected.  It should
> > +# appear at the requested line.  When the bug in GDB was present the
> > +# breakpoint would be placed on one of the following lines instead.
> > +gdb_test "info breakpoints" \
> > +    ".* in callee at \[^\r\n\]+${srcfile4}:22\\y.*" \
> > +    "check for breakpoint at ${srcfile4}"
> > +
> > +# Delete all breakpoints so that the output of "info breakpoints"
> > +# below will only contain a single breakpoint.
> > +delete_breakpoints
> > +
> > +# Place a breakpoint within the function in the header file.
> > +gdb_breakpoint "${srcfile3}:19"
> > +
> > +# Check that the breakpoint was placed where we expected.  It should
> > +# appear at the requested line.  When the bug in GDB was present the
> > +# breakpoint would be placed on one of the following lines instead.
> > +gdb_test "info breakpoints" \
> > +    ".* in callee at \[^\r\n\]+${srcfile3}:19\\y.*" \
> > +    "check for breakpoint at ${srcfile3}"
> > +
> > +# Line table entry for line 18 will have been discarded, so this
> > +# brekpoint will be at the same location as line 19.
> > +gdb_test "break ${srcfile3}:18" \
> > +    "Note: breakpoint $decimal also set at pc $hex.*"
> > diff --git a/gdb/testsuite/gdb.dwarf2/dw2-inline-header-3.exp b/gdb/testsuite/gdb.dwarf2/dw2-inline-header-3.exp
> > new file mode 100644
> > index 00000000000..c683dc4bb8a
> > --- /dev/null
> > +++ b/gdb/testsuite/gdb.dwarf2/dw2-inline-header-3.exp
> > @@ -0,0 +1,192 @@
> > +# Copyright 2020 Free Software Foundation, Inc.
> > +
> > +# This program is free software; you can redistribute it and/or modify
> > +# it under the terms of the GNU General Public License as published by
> > +# the Free Software Foundation; either version 3 of the License, or
> > +# (at your option) any later version.
> > +#
> > +# This program is distributed in the hope that it will be useful,
> > +# but WITHOUT ANY WARRANTY; without even the implied warranty of
> > +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
> > +# GNU General Public License for more details.
> > +#
> > +# You should have received a copy of the GNU General Public License
> > +# along with this program.  If not, see <http://www.gnu.org/licenses/>.
> > +
> > +# Setup a line table where:
> > +#
> > +# | Addr | File | Line | Stmt |
> > +# |------|------|------|------|
> > +# | 1    | 1    | 16   | Y    |
> > +# | 2    | 1    | 17   | Y    |
> > +# | 3    | 2    | 21   | Y    |
> > +# | 4    | 2    | 22   | Y    |
> > +# | 4    | 1    | 18   | N    |
> > +# | 5    | 1    | 19   | N    |
> > +# | 6    | 1    | 20   | Y    |
> > +# | 7    | 1    | END  | Y    |
> > +# |------|------|------|------|
> > +#
> > +# Break at file 2, line 22, then single instruction step forward.  We
> > +# should pass through line 19 and then encounter line 20.
> > +#
> > +# Currently we don't expect GDB to see file 1, line 18, as this is a
> > +# non-stmt line in a different file at the same address as the
> > +# previous is-stmt line.
> > +
> > +load_lib dwarf.exp
> > +
> > +# This test can only be run on targets which support DWARF-2 and use gas.
> > +if {![dwarf2_support]} {
> > +    return 0
> > +}
> > +
> > +# The .c files use __attribute__.
> > +if [get_compiler_info] {
> > +    return -1
> > +}
> > +if !$gcc_compiled {
> > +    return 0
> > +}
> > +
> > +standard_testfile dw2-inline-header-lbls.c dw2-inline-header.S \
> > +    dw2-inline-header.c dw2-inline-header.h
> > +
> > +set asm_file [standard_output_file $srcfile2]
> > +Dwarf::assemble $asm_file {
> > +    global srcdir subdir srcfile srcfile3 srcfile4
> > +    declare_labels lines_label callee_subprog_label
> > +
> > +    get_func_info main
> > +
> > +    cu {} {
> > +	compile_unit {
> > +	    {producer "gcc" }
> > +	    {language @DW_LANG_C}
> > +	    {name ${srcfile3}}
> > +	    {low_pc 0 addr}
> > +	    {stmt_list ${lines_label} DW_FORM_sec_offset}
> > +	} {
> > +	    callee_subprog_label: subprogram {
> > +		{external 1 flag}
> > +		{name callee}
> > +		{inline 3 data1}
> > +	    }
> > +	    subprogram {
> > +		{external 1 flag}
> > +		{name main}
> > +		{low_pc $main_start addr}
> > +		{high_pc "$main_start + $main_len" addr}
> > +	    } {
> > +		inlined_subroutine {
> > +		    {abstract_origin %$callee_subprog_label}
> > +		    {low_pc line_label_1 addr}
> > +		    {high_pc line_label_7 addr}
> > +		    {call_file 1 data1}
> > +		    {call_line 18 data1}
> > +		}
> > +	    }
> > +	}
> > +    }
> > +
> > +    lines {version 2 default_is_stmt 1} lines_label {
> > +	include_dir "${srcdir}/${subdir}"
> > +	file_name "$srcfile3" 1
> > +	file_name "$srcfile4" 1
> > +
> > +	program {
> > +	    {DW_LNE_set_address line_label_1}
> > +	    {DW_LNS_advance_line 15}
> > +	    {DW_LNS_copy}
> > +
> > +	    {DW_LNE_set_address line_label_2}
> > +	    {DW_LNS_advance_line 1}
> > +	    {DW_LNS_copy}
> > +
> > +	    {DW_LNS_set_file 2}
> > +	    {DW_LNE_set_address line_label_3}
> > +	    {DW_LNS_advance_line 4}
> > +	    {DW_LNS_copy}
> > +
> > +	    {DW_LNE_set_address line_label_4}
> > +	    {DW_LNS_advance_line 1}
> > +	    {DW_LNS_copy}
> > +
> > +	    {DW_LNS_advance_line -4}
> > +	    {DW_LNS_set_file 1}
> > +	    {DW_LNS_negate_stmt}
> > +	    {DW_LNS_copy}
> > +
> > +	    {DW_LNE_set_address line_label_5}
> > +	    {DW_LNS_advance_line 1}
> > +	    {DW_LNS_copy}
> > +
> > +	    {DW_LNE_set_address line_label_6}
> > +	    {DW_LNS_advance_line 1}
> > +	    {DW_LNS_negate_stmt}
> > +	    {DW_LNS_copy}
> > +
> > +	    {DW_LNE_set_address line_label_7}
> > +	    {DW_LNE_end_sequence}
> > +	}
> > +    }
> > +}
> > +
> > +if { [prepare_for_testing "failed to prepare" ${testfile} \
> > +	  [list $srcfile $asm_file] {nodebug optimize=-O1}] } {
> > +    return -1
> > +}
> > +
> > +if ![runto_main] {
> > +    return -1
> > +}
> > +
> > +# Delete all breakpoints so that the output of "info breakpoints"
> > +# below will only contain a single breakpoint.
> > +delete_breakpoints
> > +
> > +# Place a breakpoint within the function in the header file.
> > +gdb_breakpoint "${srcfile4}:22"
> > +
> > +# Check that the breakpoint was placed where we expected.  It should
> > +# appear at the requested line.  When the bug in GDB was present the
> > +# breakpoint would be placed on one of the following lines instead.
> > +gdb_test "info breakpoints" \
> > +    ".* in callee at \[^\r\n\]+${srcfile4}:22\\y.*"
> > +
> > +gdb_continue_to_breakpoint "${srcfile4}:22" \
> > +    ".* ${srcfile4} : 22 .*"
> > +
> > +# Now single instruction step forward.  Eventually we should hit
> > +# ${srcfile3}:20, but before we do we should hit the non-statement
> > +# line ${srcfile3}:19.
> > +#
> > +# We don't know how many instructions we'll need to step, but 100
> > +# should be enough for everyone (surely), and this stops us looping
> > +# forever if something goes wrong.
> > +set found_line_19 0
> > +set found_line_20 0
> > +set keep_going 1
> > +for { set i 0 } { $i < 100 && $keep_going } { incr i } {
> > +    set keep_going 0
> > +    gdb_test_multiple "stepi" "stepi ${i}" {
> > +	-re "${srcfile3} : 19 .*${gdb_prompt} " {
> > +	    set found_line_19 1
> > +	    set keep_going 1
> > +	}
> > +
> > +	-re "${srcfile3} : 20 .*${gdb_prompt} " {
> > +	    set found_line_20 1
> > +	}
> > +
> > +	-re "${srcfile4} : 22 .*${gdb_prompt} " {
> > +	    # Not left line 22 yet.
> > +	    set keep_going 1
> > +	}
> > +    }
> > +}
> > +
> > +gdb_assert { $found_line_19 && $found_line_20 } \
> > +    "found line 19 and 20"
> > +
> > +
> > diff --git a/gdb/testsuite/gdb.dwarf2/dw2-inline-header-lbls.c b/gdb/testsuite/gdb.dwarf2/dw2-inline-header-lbls.c
> > new file mode 100644
> > index 00000000000..a1b7b17cbeb
> > --- /dev/null
> > +++ b/gdb/testsuite/gdb.dwarf2/dw2-inline-header-lbls.c
> > @@ -0,0 +1,46 @@
> > +/* Copyright 2020 Free Software Foundation, Inc.
> > +
> > +   This program is free software; you can redistribute it and/or modify
> > +   it under the terms of the GNU General Public License as published by
> > +   the Free Software Foundation; either version 3 of the License, or
> > +   (at your option) any later version.
> > +
> > +   This program is distributed in the hope that it will be useful,
> > +   but WITHOUT ANY WARRANTY; without even the implied warranty of
> > +   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
> > +   GNU General Public License for more details.
> > +
> > +   You should have received a copy of the GNU General Public License
> > +   along with this program.  If not, see <http://www.gnu.org/licenses/>.  */
> > +
> > +/* Used to insert labels with which we can build a fake line table.  */
> > +#define LL(N) asm ("line_label_" #N ": .globl line_label_" #N)
> > +
> > +volatile int var;
> > +volatile int bar;
> > +
> > +/* Generate some code to take up some space.  */
> > +#define FILLER do { \
> > +    var = 99;	    \
> > +} while (0)
> > +
> > +int
> > +main ()
> > +{					/* main prologue */
> > +  asm ("main_label: .globl main_label");
> > +  LL (1);	// F1, Ln 16
> > +  FILLER;
> > +  LL (2);	// F1, Ln 17
> > +  FILLER;
> > +  LL (3);	// F2, Ln 21
> > +  FILLER;
> > +  LL (4);	// F2, Ln 22 // F1, Ln 18, !S
> > +  FILLER;
> > +  LL (5);	// F1, Ln 19 !S
> > +  FILLER;
> > +  LL (6);	// F1, Ln 20
> > +  FILLER;
> > +  LL (7);
> > +  FILLER;
> > +  return 0;				/* main end */
> > +}
> > diff --git a/gdb/testsuite/gdb.dwarf2/dw2-inline-header.c b/gdb/testsuite/gdb.dwarf2/dw2-inline-header.c
> > new file mode 100644
> > index 00000000000..a8331268a09
> > --- /dev/null
> > +++ b/gdb/testsuite/gdb.dwarf2/dw2-inline-header.c
> > @@ -0,0 +1,24 @@
> > +/* Copyright 2020 Free Software Foundation, Inc.
> > +
> > +   This program is free software; you can redistribute it and/or modify
> > +   it under the terms of the GNU General Public License as published by
> > +   the Free Software Foundation; either version 3 of the License, or
> > +   (at your option) any later version.
> > +
> > +   This program is distributed in the hope that it will be useful,
> > +   but WITHOUT ANY WARRANTY; without even the implied warranty of
> > +   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
> > +   GNU General Public License for more details.
> > +
> > +   You should have received a copy of the GNU General Public License
> > +   along with this program.  If not, see <http://www.gnu.org/licenses/>.  */
> > +
> > +/* dw2-inline-header.c : 16 */
> > +/* dw2-inline-header.c : 17 */
> > +/* dw2-inline-header.c : 18 */
> > +/* dw2-inline-header.c : 19 */
> > +/* dw2-inline-header.c : 20 */
> > +/* dw2-inline-header.c : 21 */
> > +/* dw2-inline-header.c : 22 */
> > +/* dw2-inline-header.c : 23 */
> > +/* dw2-inline-header.c : 24 */
> > diff --git a/gdb/testsuite/gdb.dwarf2/dw2-inline-header.h b/gdb/testsuite/gdb.dwarf2/dw2-inline-header.h
> > new file mode 100644
> > index 00000000000..7233acbcd76
> > --- /dev/null
> > +++ b/gdb/testsuite/gdb.dwarf2/dw2-inline-header.h
> > @@ -0,0 +1,24 @@
> > +/* Copyright 2020 Free Software Foundation, Inc.
> > +
> > +   This program is free software; you can redistribute it and/or modify
> > +   it under the terms of the GNU General Public License as published by
> > +   the Free Software Foundation; either version 3 of the License, or
> > +   (at your option) any later version.
> > +
> > +   This program is distributed in the hope that it will be useful,
> > +   but WITHOUT ANY WARRANTY; without even the implied warranty of
> > +   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
> > +   GNU General Public License for more details.
> > +
> > +   You should have received a copy of the GNU General Public License
> > +   along with this program.  If not, see <http://www.gnu.org/licenses/>.  */
> > +
> > +/* dw2-inline-header.h : 16 */
> > +/* dw2-inline-header.h : 17 */
> > +/* dw2-inline-header.h : 18 */
> > +/* dw2-inline-header.h : 19 */
> > +/* dw2-inline-header.h : 20 */
> > +/* dw2-inline-header.h : 21 */
> > +/* dw2-inline-header.h : 22 */
> > +/* dw2-inline-header.h : 23 */
> > +/* dw2-inline-header.h : 24 */
> > 


  parent reply	other threads:[~2020-04-04 22:23 UTC|newest]

Thread overview: 79+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-02-05 11:37 [PATCH 0/2] Line table is_stmt support Andrew Burgess
2020-02-05 11:37 ` [PATCH 2/2] gdb: Add support for tracking the DWARF line table is-stmt field Andrew Burgess
2020-02-05 17:55   ` Bernd Edlinger
2020-02-10 18:30     ` Bernd Edlinger
2020-02-11 13:57     ` Andrew Burgess
2020-02-14 20:05       ` Bernd Edlinger
2020-03-05 18:01         ` Bernd Edlinger
2020-03-08 12:50           ` [PATCHv2 0/2] Line table is_stmt support Andrew Burgess
2020-03-08 14:39             ` Bernd Edlinger
2020-03-10 23:01               ` Andrew Burgess
2020-03-11  6:50                 ` Simon Marchi
2020-03-11 11:28                   ` Andrew Burgess
2020-03-11 13:27                     ` Simon Marchi
2020-04-03 22:21             ` [PATCH 0/2] More regression fixing from is-stmt patches Andrew Burgess
2020-04-03 22:21             ` [PATCH 1/2] gdb/testsuite: Move helper function into lib/dwarf.exp Andrew Burgess
2020-04-06 20:18               ` Tom Tromey
2020-04-14 11:18                 ` Andrew Burgess
2020-04-03 22:21             ` [PATCH 2/2] gdb: Preserve is-stmt lines when switch between files Andrew Burgess
2020-04-04 18:07               ` Bernd Edlinger
2020-04-04 19:59                 ` Bernd Edlinger
2020-04-04 22:23                 ` Andrew Burgess [this message]
2020-04-05  0:04                   ` Bernd Edlinger
2020-04-05  0:47                   ` Bernd Edlinger
2020-04-05  8:55                   ` Bernd Edlinger
2020-04-11  3:52               ` Bernd Edlinger
2020-04-12 17:13                 ` Bernd Edlinger
2020-04-14 11:28                   ` Andrew Burgess
2020-04-14 11:37                     ` Bernd Edlinger
2020-04-14 11:41                       ` Bernd Edlinger
2020-04-14 13:08                       ` Andrew Burgess
2020-04-16 17:18                     ` Andrew Burgess
2020-04-22 21:13                       ` Tom Tromey
2020-04-25  7:06                         ` Bernd Edlinger
2020-04-27 10:34                           ` Andrew Burgess
2020-05-14 20:18                             ` Tom Tromey
2020-05-14 22:39                               ` Andrew Burgess
2020-05-15  3:35                                 ` Bernd Edlinger
2020-05-15 14:46                                   ` Andrew Burgess
2020-05-16  8:12                                     ` Bernd Edlinger
2020-05-17 17:26                                   ` Bernd Edlinger
2020-05-20 18:26                                   ` Andrew Burgess
2020-05-27 13:10                                     ` Andrew Burgess
2020-06-01  9:05                                       ` Andrew Burgess
2020-03-08 12:50           ` [PATCHv2 1/2] gdb/testsuite: Add is-stmt support to the DWARF compiler Andrew Burgess
2020-03-08 12:50           ` [PATCHv2 2/2] gdb: Add support for tracking the DWARF line table is-stmt field Andrew Burgess
2020-03-16 20:57             ` Tom Tromey
2020-03-16 22:37               ` Bernd Edlinger
2020-03-17 12:47               ` Tom Tromey
2020-03-17 18:23                 ` Tom Tromey
2020-03-17 18:51                   ` Bernd Edlinger
2020-03-17 18:56                   ` Andrew Burgess
2020-03-17 20:18                     ` Tom Tromey
2020-03-17 22:21                       ` Andrew Burgess
2020-03-23 17:30             ` [PATCH 0/3] Keep duplicate line table entries Andrew Burgess
2020-03-23 17:30             ` [PATCH 1/3] gdb/testsuite: Add compiler options parameter to function_range helper Andrew Burgess
2020-04-01 18:31               ` Tom Tromey
2020-03-23 17:30             ` [PATCH 2/3] gdb/testsuite: Add support for DW_LNS_set_file to DWARF compiler Andrew Burgess
2020-04-01 18:32               ` Tom Tromey
2020-03-23 17:30             ` [PATCH 3/3] gdb: Don't remove duplicate entries from the line table Andrew Burgess
2020-04-01 18:34               ` Tom Tromey
2020-06-01 13:26           ` [PATCH 2/2] gdb: Add support for tracking the DWARF line table is-stmt field Pedro Alves
2020-02-06  9:01   ` Luis Machado
2020-02-11 15:39     ` Andrew Burgess
2020-02-09 21:07   ` [PATCH] Fix range end handling of inlined subroutines Bernd Edlinger
2020-02-10 21:48     ` Andrew Burgess
2020-02-22  6:39     ` [PATCHv2] " Bernd Edlinger
2020-03-08 14:57       ` [PATCHv3] " Bernd Edlinger
2020-03-11 22:02         ` Andrew Burgess
2020-03-12 18:21           ` Bernd Edlinger
2020-03-12 18:27             ` Christian Biesinger
2020-03-13  8:03               ` Bernd Edlinger
2020-03-17 22:27                 ` Andrew Burgess
2020-03-19  1:33                   ` Bernd Edlinger
2020-03-21 20:31                     ` Bernd Edlinger
2020-03-23 17:53                       ` Andrew Burgess
2020-03-23 20:58                         ` Bernd Edlinger
2020-06-01 14:28                           ` Pedro Alves
2020-03-13 12:47         ` [PATCHv4] " Bernd Edlinger
2020-02-05 11:37 ` [PATCH 1/2] gdb/testsuite: Add is-stmt support to the DWARF compiler Andrew Burgess

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20200404222326.GB3917@embecosm.com \
    --to=andrew.burgess@embecosm.com \
    --cc=bernd.edlinger@hotmail.de \
    --cc=gdb-patches@sourceware.org \
    --cc=tom@tromey.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox