From: Andrew Burgess <aburgess@redhat.com>
To: Tom Tromey <tom@tromey.com>, gdb-patches@sourceware.org
Cc: Tom Tromey <tom@tromey.com>
Subject: Re: [PATCH] Fix crash in Fortran code
Date: Thu, 24 Feb 2022 19:03:26 +0000 [thread overview]
Message-ID: <875yp4axz5.fsf@redhat.com> (raw)
In-Reply-To: <20220224160418.4181081-1-tom@tromey.com>
Tom Tromey <tom@tromey.com> writes:
> PR fortran/28801 points out a gdb crash that can be provoked by
> certain Fortran code. The bug is that f77_get_upperbound assumes the
> property is either a constant or undefined, but in this case it is
> PROP_LOCEXPR.
>
> This patch fixes the crash by making this function (and the
> lower-bound one as well) do the correct check before calling
> 'const_val'.
LGTM. Great test case!
> Thanks to Andrew for writing the test case.
Thanks,
Andrew
>
> Co-authored-by: Andrew Burgess <aburgess@redhat.com>
> Bug: https://sourceware.org/bugzilla/show_bug.cgi?id=28801
> ---
> gdb/f-valprint.c | 4 +-
> gdb/testsuite/gdb.dwarf2/fortran-var-string.c | 31 ++++
> .../gdb.dwarf2/fortran-var-string.exp | 175 ++++++++++++++++++
> 3 files changed, 208 insertions(+), 2 deletions(-)
> create mode 100644 gdb/testsuite/gdb.dwarf2/fortran-var-string.c
> create mode 100644 gdb/testsuite/gdb.dwarf2/fortran-var-string.exp
>
> diff --git a/gdb/f-valprint.c b/gdb/f-valprint.c
> index 6a199f17c1e..b64750bac7b 100644
> --- a/gdb/f-valprint.c
> +++ b/gdb/f-valprint.c
> @@ -43,7 +43,7 @@ static void f77_get_dynamic_length_of_aggregate (struct type *);
> LONGEST
> f77_get_lowerbound (struct type *type)
> {
> - if (type->bounds ()->low.kind () == PROP_UNDEFINED)
> + if (type->bounds ()->low.kind () != PROP_CONST)
> error (_("Lower bound may not be '*' in F77"));
>
> return type->bounds ()->low.const_val ();
> @@ -52,7 +52,7 @@ f77_get_lowerbound (struct type *type)
> LONGEST
> f77_get_upperbound (struct type *type)
> {
> - if (type->bounds ()->high.kind () == PROP_UNDEFINED)
> + if (type->bounds ()->high.kind () != PROP_CONST)
> {
> /* We have an assumed size array on our hands. Assume that
> upper_bound == lower_bound so that we show at least 1 element.
> diff --git a/gdb/testsuite/gdb.dwarf2/fortran-var-string.c b/gdb/testsuite/gdb.dwarf2/fortran-var-string.c
> new file mode 100644
> index 00000000000..a677b7e6550
> --- /dev/null
> +++ b/gdb/testsuite/gdb.dwarf2/fortran-var-string.c
> @@ -0,0 +1,31 @@
> +/* Copyright 2022 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/>. */
> +
> +/* In the generated DWARF, we'll pretend that ARG is a string with dynamic
> + length. */
> +int
> +main_helper (void *arg)
> +{
> + asm ("main_helper_label: .globl main_helper_label");
> + return 0;
> +}
> +
> +int
> +main (void)
> +{
> + asm ("main_label: .globl main_label");
> + main_helper (0);
> + return 0;
> +}
> diff --git a/gdb/testsuite/gdb.dwarf2/fortran-var-string.exp b/gdb/testsuite/gdb.dwarf2/fortran-var-string.exp
> new file mode 100644
> index 00000000000..793c4418fd6
> --- /dev/null
> +++ b/gdb/testsuite/gdb.dwarf2/fortran-var-string.exp
> @@ -0,0 +1,175 @@
> +# Copyright 2022 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/>.
> +load_lib dwarf.exp
> +
> +# This test can only be run on targets which support DWARF-2 and use gas.
> +if {![dwarf2_support]} {
> + return 0
> +}
> +
> +standard_testfile .c -dw.S
> +
> +# We need to know the size of integer and address types in order
> +# to write some of the debugging info we'd like to generate.
> +#
> +# For that, we ask GDB by debugging our dynarr-ptr.c program.
> +# Any program would do, but since we already have dynarr-ptr.c
> +# specifically for this testcase, might as well use that.
> +
> +if { [prepare_for_testing "failed to prepare" ${testfile} ${srcfile}] } {
> + return -1
> +}
> +
> +# Make some DWARF for the test.
> +set asm_file [standard_output_file $srcfile2]
> +Dwarf::assemble $asm_file {
> + set int_size [get_sizeof "int" 4]
> +
> + get_func_info main
> + get_func_info main_helper
> +
> + cu {} {
> + DW_TAG_compile_unit {
> + {DW_AT_language @DW_LANG_Fortran90}
> + {DW_AT_name fortran-var-string.f90}
> + {DW_AT_comp_dir /tmp}
> + } {
> + declare_labels integer_label string_label array_lb_label \
> + array_ub_label
> +
> + DW_TAG_subprogram {
> + {name main}
> + {low_pc $main_helper_start addr}
> + {high_pc $main_helper_len data8}
> + {DW_AT_type :$integer_label}
> + {DW_AT_decl_file 1 data1}
> + {DW_AT_decl_line 1 data1}
> + }
> +
> + DW_TAG_subprogram {
> + {name test_1_func}
> + {low_pc $main_start addr}
> + {high_pc $main_len data8}
> + {DW_AT_type :$integer_label}
> + {DW_AT_decl_file 1 data1}
> + {DW_AT_decl_line 2 data1}
> + } {
> + formal_parameter {
> + {name arg1}
> + {type :$string_label}
> + }
> + }
> +
> + DW_TAG_subprogram {
> + {name test_2_func}
> + {low_pc $main_start addr}
> + {high_pc $main_len data8}
> + {DW_AT_type :$integer_label}
> + {DW_AT_decl_file 1 data1}
> + {DW_AT_decl_line 3 data1}
> + } {
> + formal_parameter {
> + {name arg1}
> + {type :$array_ub_label}
> + }
> + }
> +
> + DW_TAG_subprogram {
> + {name test_3_func}
> + {low_pc $main_start addr}
> + {high_pc $main_len data8}
> + {DW_AT_type :$integer_label}
> + {DW_AT_decl_file 1 data1}
> + {DW_AT_decl_line 4 data1}
> + } {
> + formal_parameter {
> + {name arg1}
> + {type :$array_lb_label}
> + }
> + }
> +
> + integer_label: DW_TAG_base_type {
> + {DW_AT_byte_size $int_size DW_FORM_sdata}
> + {DW_AT_encoding @DW_ATE_signed}
> + {DW_AT_name integer}
> + }
> +
> + string_label: DW_TAG_string_type {
> + {DW_AT_byte_size $int_size DW_FORM_sdata}
> + {DW_AT_name .str.arg}
> + {DW_AT_string_length {} DW_FORM_block1}
> + }
> +
> + array_lb_label: DW_TAG_array_type {
> + {DW_AT_ordering 1 data1}
> + {DW_AT_type :$integer_label}
> + } {
> + DW_TAG_subrange_type {
> + {DW_AT_lower_bound {} DW_FORM_block1}
> + {DW_AT_upper_bound 10 DW_FORM_data1}
> + }
> + }
> +
> + array_ub_label: DW_TAG_array_type {
> + {DW_AT_ordering 1 data1}
> + {DW_AT_type :$integer_label}
> + } {
> + DW_TAG_subrange_type {
> + {DW_AT_upper_bound {} DW_FORM_block1}
> + }
> + }
> + }
> + }
> +}
> +
> +# Now that we've generated the DWARF debugging info, rebuild our
> +# program using our debug info instead of the info generated by
> +# the compiler.
> +
> +if { [prepare_for_testing "failed to prepare" ${testfile} \
> + [list $srcfile $asm_file] {nodebug}] } {
> + return -1
> +}
> +
> +if ![runto_main] {
> + return -1
> +}
> +
> +gdb_test_no_output "set language fortran"
> +
> +gdb_test "info functions test_1_func" \
> + "2:\\s+integer test_1_func\\(character\\*\\(\\*\\)\\);"
> +
> +# We print `1` here as the bound because GDB treats this as an assumed
> +# size array, and just reports the lower bound value for the upper
> +# bound.
> +#
> +# We might, in the future, decide that there's a better way we could
> +# tell the user about the type of this array argument, when that
> +# happens it's OK to change the expected results here.
> +gdb_test "info functions test_2_func" \
> + "3:\\s+integer test_2_func\\(integer \\(1\\)\\);"
> +
> +# It's not completely clear that this error is correct here. Why
> +# can't the lower bound be a dynamic expression?
> +#
> +# This test was initially added to guard against the case where GDB
> +# was crashing if/when it saw this situation.
> +#
> +# If later on, GDB's handling of array types with a dynamic loewr
> +# bound changes, then it is possible that the expected result here
> +# should change.
> +gdb_test "info functions test_3_func" \
> + "4:\\s+Lower bound may not be '\\*' in F77"
> --
> 2.31.1
prev parent reply other threads:[~2022-02-24 19:03 UTC|newest]
Thread overview: 2+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-02-24 16:04 Tom Tromey
2022-02-24 19:03 ` Andrew Burgess [this message]
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=875yp4axz5.fsf@redhat.com \
--to=aburgess@redhat.com \
--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;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).