RE: [PATCH v3] Fix reverse stepping multiple contiguous PC ranges over the line table.

[Carl Love <cel@us.ibm.com>]

Bruno:

Thanks for the review.  I addressed your comments as mentioned below.

I will post version 4 with the changes.

FYI, I will be out of the office from May 11 thru May 15.  So will
reply to any additional comments when I return.

On Wed, 2023-05-10 at 15:47 +0200, Bruno Larsen wrote:
> On 04/05/2023 17:59, Carl Love wrote:
> > Bruno, GDB maintainers:
> > 
> > Version 3, added the gcc version check as discussed further from
> > version 2 of the patch.  Also updated the tests to check for
> > supporting
> > 

<snip>

> >   extern struct symtab *find_pc_line_symtab (CORE_ADDR);
> > diff --git a/gdb/testsuite/gdb.reverse/func-map-to-same-line.c
> > b/gdb/testsuite/gdb.reverse/func-map-to-same-line.c
> > new file mode 100644
> > index 00000000000..412ab180943
> > --- /dev/null
> > +++ b/gdb/testsuite/gdb.reverse/func-map-to-same-line.c
> > @@ -0,0 +1,36 @@
> > +/* Copyright 2008-2023 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 L

<snip>

> > +
> > +require supports_reverse
> > +
> > +# This test uses the gcc no-column-info command which was added in
> > gcc 7.1.
> > +require get_compiler_info  "gcc-7-*"

I put the compiler check in last.  When I ran it, I obviously didn't
double check gdb/testsuite/gdb.log to make sure it really worked.  I
normally try to make a point of double checking the log file.  I have
been burned before thinking it was OK when there were no errors visible
on the command line. The above command fails if you check the log file.
> 
> By constructing your regex like this, you are only allowing this test
> to 
> be run on gcc 7. Anything later is also not accepted. I would do 
> something like (Warning, untested)
> 
> require get_compiler_info "gcc"
> require !get_compiler_info "gcc-[1-6]-*"

I couldn't get require to work like that.  The get_compiler_info
doesn't seem take "gcc" as an argument.  

I changed the test to:

if {![test_compiler_info {gcc-*}]
    || [test_compiler_info {gcc-[1-6]-*}]} {
    return
}

With this, I do see the correct number of passes in
gdb/testsuite/gdb.log.

> 
> Which requires gcc, but does not allow versions 1 to 6. There is 
> probably a way to do it with a single require line, but I'm not the
> best 
> with regexes.
> 
> > +
> > +proc run_tests {msg} {
> > +    global srcfile
> > +    global executable
> > +
> > +    runto_main
> > +    set target_remote [gdb_is_target_remote]
> 
> When probing for target remote, GDB will emit pass/fails with
> hardcoded 
> names, so the current proc setup gives us some duplicated test names.
> 
> I would suggest that, instead of passing a message as a parameter,
> you 
> wrapped all function calls in a with_test_prefix scope, like:
> 
> with_test_prefix "with-column-info" {
>      run_test
> }

OK, I changed from passing in an argument and did the wrapped calls to
run_test instead. Note, this still didn't fix the duplicate test names
for turning on record.

> 
> > +
> > +    gdb_test_no_output "record" "turn on process record"
> > +
> > +    # This regression test verifies the reverse-step and reverse-
> > next commands
> > +    # work properly when executing backwards thru a source line
> > containing
> > +    # two function calls on the same source line, i.e. func1 ();
> > func2 ();
> > +    # This test is compiled so the dwarf info not contain the line
> > table
> > +    # information.
> > +
> > +    # Test 1, reverse-next command
> > +    # Set breakpoint at the line after the function calls.
> > +    set bp_start_reverse_test [gdb_get_line_number "START REVERSE
> > TEST" \
> > +				   $srcfile]
> > +    gdb_breakpoint $srcfile:$bp_start_reverse_test temporary
> > +
> > +    # Continue to break point for reverse-next test.
> > +    # Command definition:  reverse-next [count]
> > +    #   Run backward to the beginning of the previous line
> > executed in the
> > +    #   current (innermost) stack frame. If the line contains
> > function calls,
> > +    #   they will be “un-executed” without stopping. Starting from
> > the first
> > +    #   line of a function, reverse-next will take you back to the
> > caller of
> > +    #   that function, before the function was called, just as the
> > normal next
> > +    #   command would take you from the last line of a function
> > back to its
> > +    #   return to its caller 2 .
> > +    gdb_continue_to_breakpoint \
> > +	"$msg: test1: stopped at command reverse-next test start
> > location" \
> > +	".*$srcfile:$bp_start_reverse_test\r\n.*"
> > +
> > +    # The reverse-next should step all the way back to the
> > beginning of the
> > +    # line, i.e. at the beginning of the func1 call.
> > +    gdb_test "reverse-next" ".*func1 \\(\\); func2 \\(\\);.*" \
> > +	"$msg: test1: reverse-next to line with two functions"
> > +
> > +    # We should be stopped at the first instruction of the line. A
> > reverse-step
> > +    # should step back and stop at the beginning of the previous
> > line b = 2,
> > +    # i.e. not in func1 ().
> > +    gdb_test "reverse-stepi" ".*b = 2;.*" \
> > +	"$msg: test1: reverse-stepi to previous line b = 2"
> > +
> > +
> > +    # Setup for test 2
> > +    clean_restart $executable
> > +    runto_main
> > +
> > +    gdb_test_no_output "record" "turn on process record"
> 
> This gives a duplicate test name from setting up for the first test. 
> Adding "test 2:(...)" in here solves it.

I couldn't figure out how to get the above syntax to work.  So I used
the with_test_prefix instead, i.e.

    with_test_prefix "test2" {
        gdb_test_no_output "record" "turn on process record"
    }

That fixed the duplicate names.  I also wrapped the first record with
"test1" for consistency.

> 
> > +
> > +    # Test 2, reverse-step command
> > +    # Set breakpoint at the line after the function calls.
> > +    gdb_breakpoint $srcfile:$bp_start_reverse_test temporary
> > +
> > 

<snip>

> > +# the same line to do the needed testing.
> > +
> > +# This test can only be run on targets which support DWARF-2 and
> > use gas.
> > +require dwarf2_support
> > +load_lib dwarf.exp
> the library has to be imported before the "require"

OK, switched the order of the lines.

> > +
> > +# The DWARF assembler requires the gcc compiler.
> > +require is_c_compiler_gcc
> > +
> > +# This test suitable only for process that can do reverse
> > execution
> > +requires supports_reverse
> 
> s/requires/require

Fixed.
            
                   Carl