[Bug gdb/27340] New: Output does not respect set width value in some cases.

[Bug gdb/27340] New: Output does not respect set width value in some cases.

[abidh at codesourcery dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=27340

            Bug ID: 27340
           Summary: Output does not respect set width value in some cases.
           Product: gdb
           Version: HEAD
            Status: UNCONFIRMED
          Severity: normal
          Priority: P2
         Component: gdb
          Assignee: unassigned at sourceware dot org
          Reporter: abidh at codesourcery dot com
  Target Milestone: ---

I was investigating test failure of gdb.base/style.exp and noticed a problem
that can be seen in the following example session below.

(gdb) frame
#0  main (argc=1, argv=0x7fffffffdf08) at
/home/abidh/work/gdb/src/upstream/gdb/testsuite/gdb.base/style.c:45
45        return some_called_function (); /* break here */
(gdb) set width 20
(gdb) frame
#0  main (argc=1, 
    argv=0x7fffffffdf08)
    at /home/abidh/work/gdb/src/upstream/gdb/testsuite/gdb.base/style.c:45
45        return some_called_function (); /* break here */
(gdb) set width 24
(gdb) frame
#0  main (argc=1, 
    argv=0x7fffffffdf08) at
/home/abidh/work/gdb/src/upstream/gdb/testsuite/gdb.base/style.c:45
45        return some_called_function (); /* break here */
(gdb) 

Note that:
Filename is printed on the next line with width == 20
Filename is printed on the same line when width == 24

Here 24 is significant. Problem only manifests itself when number of digits in
the argv are (width - 12). The gdb.base/style.exp checks with width of 20 and
30. As it happens, the 'argv' parameter had 8 digit for aarch64-none-elf target
so the test failed on this platform for me.

I stepped through code a bit. When 'fputs_maybe_filtered' is printing stuff and
output is about to cross width, it will set 'chars_printed' to 0 but as
'wrap_column' is 0, it will not add a newline in the output. Now if we end up
calling 'wrap_here' before printing any more character then 'wrap_column' will
stay 0 as 'chars_printed' is still 0.

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug gdb/27340] Output does not respect set width value in some cases.

[tromey at sourceware dot org]

https://sourceware.org/bugzilla/show_bug.cgi?id=27340

Tom Tromey <tromey at sourceware dot org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |tromey at sourceware dot org

--- Comment #1 from Tom Tromey <tromey at sourceware dot org> ---
Does this still happen?

I wonder if it's possible to update the test to show the failure.
Or copy it and write a new test.

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug gdb/27340] Output does not respect set width value in some cases.

[abidh at sourceware dot org]

https://sourceware.org/bugzilla/show_bug.cgi?id=27340

--- Comment #2 from Hafiz Abid Qadeer <abidh at sourceware dot org> ---
I think problem is still there. Here is the small session that shows it. Note
that file is printed on same line only for width=24 and not for 23 or 25.
Following patch can make the style.exp fail on my x86_64 ubuntu 20.04 machine.

diff --git a/gdb/testsuite/gdb.base/style.exp
b/gdb/testsuite/gdb.base/style.exp
index 2cd155d2cf2..4c89bfb55bc 100644
--- a/gdb/testsuite/gdb.base/style.exp
+++ b/gdb/testsuite/gdb.base/style.exp
@@ -131,7 +131,7 @@ proc run_style_tests { } {
        # Regression test for a bug where line-wrapping would occur at
        # the wrong spot with styling.  There were different bugs at
        # different widths, so try two.
-       foreach width {20 30} {
+       foreach width {24 30} {
            set argv_len [string length $argv]
            if { $argv_len == 0 } {
                continue
@@ -159,7 +159,7 @@ proc run_style_tests { } {
            # - closing parenthesis
            set line_len [expr 4 + $argv_len + 1]

-           if { $line_len > $width } {
+           if { $line_len >= $width } {
                # At on the next line.
                set re_styled $re1_styled
            } else {


gdb  gdb/testsuite/outputs/gdb.base/style/style
GNU gdb (GDB) 12.0.50.20220104-git
...
Reading symbols from gdb/testsuite/outputs/gdb.base/style/style...
(gdb) start
Temporary breakpoint 1 at 0x65b: file
/home/abidh/work/gdb/src/upstream/gdb/testsuite/gdb.base/style.c, line 52.
Starting program:
/mnt/work/obj/gdb/upstream/native/obj/gdb/testsuite/outputs/gdb.base/style/style 

Temporary breakpoint 1, main (argc=1, argv=0x7fffffffdef8) at
/home/abidh/work/gdb/src/upstream/gdb/testsuite/gdb.base/style.c:52
52        return some_called_function (); /* break here */
(gdb) set width 23
(gdb) frame
#0  main (argc=1, 
    argv=0x7fffffffdef8)
    at /home/abidh/work/gdb/src/upstream/gdb/testsuite/gdb.base/style.c:52
52        return some_called_function (); /* break here */
(gdb) set width 24
(gdb) frame
#0  main (argc=1, 
    argv=0x7fffffffdef8) at
/home/abidh/work/gdb/src/upstream/gdb/testsuite/gdb.base/style.c:52
52        return some_called_function (); /* break here */
(gdb) set width 25
(gdb) frame
#0  main (argc=1, 
    argv=0x7fffffffdef8)
    at /home/abidh/work/gdb/src/upstream/gdb/testsuite/gdb.base/style.c:52
52        return some_called_function (); /* break here */

-- 
You are receiving this mail because:
You are on the CC list for the bug.