* [PATCH v3] Dwarf: Don't add nameless modules to partial symbol table.
@ 2019-01-05 11:44 Pawel Wodkowski
2019-02-20 11:50 ` [PING] " Paweł Wódkowski
2019-02-27 14:18 ` [PATCH v4] " Tomasz Kulasek
0 siblings, 2 replies; 5+ messages in thread
From: Pawel Wodkowski @ 2019-01-05 11:44 UTC (permalink / raw)
To: gdb-patches, andrew.burgess
From: Bernhard Heckel <bernhard.heckel@intel.com>
A name for BLOCK DATA in Fortran is optional. If no
name has been assigned, GDB will crash during read-in of DWARF
when BLOCK DATA is represented via DW_TAG_module.
BLOCK DATA is used for one-time initialization of
non-pointer variables in named common blocks.
As of now there is no issue when gfortran is used as
DW_TAG_module is not emited. For Intel ifort the nameless
DW_TAG_module is present and have following form:
...
<1><dd>: Abbrev Number: 7 (DW_TAG_module)
<de> DW_AT_decl_line : 46
<df> DW_AT_decl_file : 1
<e0> DW_AT_description : (indirect string, offset: 0x110): block
data
<e4> DW_AT_high_pc : 0x402bb7
<ec> DW_AT_low_pc : 0x402bb7
...
Missing name lead to crash in add_partial_symbol function
during length calculation.
2016-06-15 Bernhard Heckel <bernhard.heckel@intel.com>
gdb/Changelog:
* dwarf2read.c (add_partial_symbol): Skip nameless modules.
gdb/testsuite/Changelog:
* gdb.fortran/block-data.f: New.
* gdb.fortran/block-data.exp: New.
Changes from V1 to V2:
* No changes - just sent with other patches in series.
Changes from V2 to V3:
* Decouple from other patches and send stand alone.
* Provide example DW_TAG_module in commit message.
* Fix mistakes in comments.
---
gdb/dwarf2read.c | 13 +++++---
gdb/testsuite/gdb.fortran/block-data.exp | 49 ++++++++++++++++++++++++++++
gdb/testsuite/gdb.fortran/block-data.f | 56 ++++++++++++++++++++++++++++++++
3 files changed, 113 insertions(+), 5 deletions(-)
create mode 100644 gdb/testsuite/gdb.fortran/block-data.exp
create mode 100644 gdb/testsuite/gdb.fortran/block-data.f
diff --git a/gdb/dwarf2read.c b/gdb/dwarf2read.c
index d2a8cd44f9a5..389b21d27f3f 100644
--- a/gdb/dwarf2read.c
+++ b/gdb/dwarf2read.c
@@ -8995,11 +8995,14 @@ add_partial_symbol (struct partial_die_info *pdi, struct dwarf2_cu *cu)
0, cu->language, objfile);
break;
case DW_TAG_module:
- add_psymbol_to_list (actual_name, strlen (actual_name),
- built_actual_name != NULL,
- MODULE_DOMAIN, LOC_TYPEDEF, -1,
- &objfile->global_psymbols,
- 0, cu->language, objfile);
+ /* In Fortran 77 there might be a "BLOCK DATA" module available wihout
+ any name. If so, we skip the module as it doesn't bring any value. */
+ if (actual_name != nullptr)
+ add_psymbol_to_list (actual_name, strlen (actual_name),
+ built_actual_name != nullptr,
+ MODULE_DOMAIN, LOC_TYPEDEF, -1,
+ &objfile->global_psymbols,
+ 0, cu->language, objfile);
break;
case DW_TAG_class_type:
case DW_TAG_interface_type:
diff --git a/gdb/testsuite/gdb.fortran/block-data.exp b/gdb/testsuite/gdb.fortran/block-data.exp
new file mode 100644
index 000000000000..bb40be23d26b
--- /dev/null
+++ b/gdb/testsuite/gdb.fortran/block-data.exp
@@ -0,0 +1,49 @@
+# Copyright 2019 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/>.
+
+# This test is supposed to test anonymous block-data statement.
+
+if { [skip_fortran_tests] } { return -1 }
+
+standard_testfile .f
+load_lib "fortran.exp"
+
+if {[prepare_for_testing $testfile.exp $testfile $srcfile {debug f90}]} {
+ return -1
+}
+
+if ![runto MAIN__] then {
+ untested "couldn't run to breakpoint MAIN__"
+ return -1
+}
+
+gdb_test "print doub1" "= 1.11\\d+" "print doub1, default values"
+gdb_test "print doub2" "= 2.22\\d+" "print doub2, default values"
+gdb_test "print char1" "= 'abcdef'" "print char1, default values"
+gdb_test "print char2" "= 'ghijkl'" "print char2, default values"
+
+gdb_breakpoint [gdb_get_line_number "! BP_BEFORE_SUB"]
+gdb_continue_to_breakpoint "! BP_BEFORE_SUB" ".*! BP_BEFORE_SUB.*"
+gdb_test "print doub1" "= 11.11\\d+" "print doub1, before sub"
+gdb_test "print doub2" "= 22.22\\d+" "print doub2, before sub"
+gdb_test "print char1" "= 'ABCDEF'" "print char1, before sub"
+gdb_test "print char2" "= 'GHIJKL'" "print char2, before sub"
+
+gdb_breakpoint [gdb_get_line_number "! BP_SUB"]
+gdb_continue_to_breakpoint "! BP_SUB" ".*! BP_SUB.*"
+gdb_test "print doub1" "= 11.11\\d+" "print char1, in sub"
+gdb_test "print doub2" "= 22.22\\d+" "print doub2, in sub"
+gdb_test "print char1" "= 'ABCDEF'" "print char1, in sub"
+gdb_test "print char2" "= 'GHIJKL'" "print char2, in sub"
diff --git a/gdb/testsuite/gdb.fortran/block-data.f b/gdb/testsuite/gdb.fortran/block-data.f
new file mode 100644
index 000000000000..5f50e644d303
--- /dev/null
+++ b/gdb/testsuite/gdb.fortran/block-data.f
@@ -0,0 +1,56 @@
+! Copyright 2019 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/>.
+!
+! Test if GDB can handle block data without global name
+!
+! MAIN
+ PROGRAM bdata
+ DOUBLE PRECISION doub1, doub2
+ CHARACTER*6 char1, char2
+
+ COMMON /BLK1/ doub1, char1
+ COMMON /BLK2/ doub2, char2
+
+ doub1 = 11.111
+ doub2 = 22.222
+ char1 = 'ABCDEF'
+ char2 = 'GHIJKL'
+ CALL sub_block_data ! BP_BEFORE_SUB
+ STOP
+ END
+
+! BLOCK DATA
+ BLOCK DATA
+
+ DOUBLE PRECISION doub1, doub2
+ CHARACTER*6 char1, char2
+
+ COMMON /BLK1/ doub1, char1
+ COMMON /BLK2/ doub2, char2
+ DATA doub1, doub2 /1.111, 2.222/
+ DATA char1, char2 /'abcdef', 'ghijkl'/
+ END
+
+! SUBROUTINE
+ SUBROUTINE sub_block_data
+
+ DOUBLE PRECISION doub1, doub2
+ CHARACTER*6 char1, char2
+
+ COMMON /BLK1/ doub1, char1
+ COMMON /BLK2/ doub2, char2
+
+ char1 = char2; ! BP_SUB
+ END
--
2.7.4
^ permalink raw reply [flat|nested] 5+ messages in thread
* [PING] Re: [PATCH v3] Dwarf: Don't add nameless modules to partial symbol table.
2019-01-05 11:44 [PATCH v3] Dwarf: Don't add nameless modules to partial symbol table Pawel Wodkowski
@ 2019-02-20 11:50 ` Paweł Wódkowski
2019-02-23 12:58 ` Joel Brobecker
2019-02-27 14:18 ` [PATCH v4] " Tomasz Kulasek
1 sibling, 1 reply; 5+ messages in thread
From: Paweł Wódkowski @ 2019-02-20 11:50 UTC (permalink / raw)
To: gdb-patches
Hi,
Can someone take a look at this patch?
Thanks
Pawel
On 05.01.2019 12:43, Pawel Wodkowski wrote:
> From: Bernhard Heckel <bernhard.heckel@intel.com>
>
> A name for BLOCK DATA in Fortran is optional. If no
> name has been assigned, GDB will crash during read-in of DWARF
> when BLOCK DATA is represented via DW_TAG_module.
> BLOCK DATA is used for one-time initialization of
> non-pointer variables in named common blocks.
>
> As of now there is no issue when gfortran is used as
> DW_TAG_module is not emited. For Intel ifort the nameless
> DW_TAG_module is present and have following form:
>
> ...
> <1><dd>: Abbrev Number: 7 (DW_TAG_module)
> <de> DW_AT_decl_line : 46
> <df> DW_AT_decl_file : 1
> <e0> DW_AT_description : (indirect string, offset: 0x110): block
> data
> <e4> DW_AT_high_pc : 0x402bb7
> <ec> DW_AT_low_pc : 0x402bb7
> ...
>
> Missing name lead to crash in add_partial_symbol function
> during length calculation.
>
> 2016-06-15 Bernhard Heckel <bernhard.heckel@intel.com>
>
> gdb/Changelog:
> * dwarf2read.c (add_partial_symbol): Skip nameless modules.
>
> gdb/testsuite/Changelog:
> * gdb.fortran/block-data.f: New.
> * gdb.fortran/block-data.exp: New.
>
> Changes from V1 to V2:
> * No changes - just sent with other patches in series.
>
> Changes from V2 to V3:
> * Decouple from other patches and send stand alone.
> * Provide example DW_TAG_module in commit message.
> * Fix mistakes in comments.
> ---
> gdb/dwarf2read.c | 13 +++++---
> gdb/testsuite/gdb.fortran/block-data.exp | 49 ++++++++++++++++++++++++++++
> gdb/testsuite/gdb.fortran/block-data.f | 56 ++++++++++++++++++++++++++++++++
> 3 files changed, 113 insertions(+), 5 deletions(-)
> create mode 100644 gdb/testsuite/gdb.fortran/block-data.exp
> create mode 100644 gdb/testsuite/gdb.fortran/block-data.f
>
> diff --git a/gdb/dwarf2read.c b/gdb/dwarf2read.c
> index d2a8cd44f9a5..389b21d27f3f 100644
> --- a/gdb/dwarf2read.c
> +++ b/gdb/dwarf2read.c
> @@ -8995,11 +8995,14 @@ add_partial_symbol (struct partial_die_info *pdi, struct dwarf2_cu *cu)
> 0, cu->language, objfile);
> break;
> case DW_TAG_module:
> - add_psymbol_to_list (actual_name, strlen (actual_name),
> - built_actual_name != NULL,
> - MODULE_DOMAIN, LOC_TYPEDEF, -1,
> - &objfile->global_psymbols,
> - 0, cu->language, objfile);
> + /* In Fortran 77 there might be a "BLOCK DATA" module available wihout
> + any name. If so, we skip the module as it doesn't bring any value. */
> + if (actual_name != nullptr)
> + add_psymbol_to_list (actual_name, strlen (actual_name),
> + built_actual_name != nullptr,
> + MODULE_DOMAIN, LOC_TYPEDEF, -1,
> + &objfile->global_psymbols,
> + 0, cu->language, objfile);
> break;
> case DW_TAG_class_type:
> case DW_TAG_interface_type:
> diff --git a/gdb/testsuite/gdb.fortran/block-data.exp b/gdb/testsuite/gdb.fortran/block-data.exp
> new file mode 100644
> index 000000000000..bb40be23d26b
> --- /dev/null
> +++ b/gdb/testsuite/gdb.fortran/block-data.exp
> @@ -0,0 +1,49 @@
> +# Copyright 2019 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/>.
> +
> +# This test is supposed to test anonymous block-data statement.
> +
> +if { [skip_fortran_tests] } { return -1 }
> +
> +standard_testfile .f
> +load_lib "fortran.exp"
> +
> +if {[prepare_for_testing $testfile.exp $testfile $srcfile {debug f90}]} {
> + return -1
> +}
> +
> +if ![runto MAIN__] then {
> + untested "couldn't run to breakpoint MAIN__"
> + return -1
> +}
> +
> +gdb_test "print doub1" "= 1.11\\d+" "print doub1, default values"
> +gdb_test "print doub2" "= 2.22\\d+" "print doub2, default values"
> +gdb_test "print char1" "= 'abcdef'" "print char1, default values"
> +gdb_test "print char2" "= 'ghijkl'" "print char2, default values"
> +
> +gdb_breakpoint [gdb_get_line_number "! BP_BEFORE_SUB"]
> +gdb_continue_to_breakpoint "! BP_BEFORE_SUB" ".*! BP_BEFORE_SUB.*"
> +gdb_test "print doub1" "= 11.11\\d+" "print doub1, before sub"
> +gdb_test "print doub2" "= 22.22\\d+" "print doub2, before sub"
> +gdb_test "print char1" "= 'ABCDEF'" "print char1, before sub"
> +gdb_test "print char2" "= 'GHIJKL'" "print char2, before sub"
> +
> +gdb_breakpoint [gdb_get_line_number "! BP_SUB"]
> +gdb_continue_to_breakpoint "! BP_SUB" ".*! BP_SUB.*"
> +gdb_test "print doub1" "= 11.11\\d+" "print char1, in sub"
> +gdb_test "print doub2" "= 22.22\\d+" "print doub2, in sub"
> +gdb_test "print char1" "= 'ABCDEF'" "print char1, in sub"
> +gdb_test "print char2" "= 'GHIJKL'" "print char2, in sub"
> diff --git a/gdb/testsuite/gdb.fortran/block-data.f b/gdb/testsuite/gdb.fortran/block-data.f
> new file mode 100644
> index 000000000000..5f50e644d303
> --- /dev/null
> +++ b/gdb/testsuite/gdb.fortran/block-data.f
> @@ -0,0 +1,56 @@
> +! Copyright 2019 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/>.
> +!
> +! Test if GDB can handle block data without global name
> +!
> +! MAIN
> + PROGRAM bdata
> + DOUBLE PRECISION doub1, doub2
> + CHARACTER*6 char1, char2
> +
> + COMMON /BLK1/ doub1, char1
> + COMMON /BLK2/ doub2, char2
> +
> + doub1 = 11.111
> + doub2 = 22.222
> + char1 = 'ABCDEF'
> + char2 = 'GHIJKL'
> + CALL sub_block_data ! BP_BEFORE_SUB
> + STOP
> + END
> +
> +! BLOCK DATA
> + BLOCK DATA
> +
> + DOUBLE PRECISION doub1, doub2
> + CHARACTER*6 char1, char2
> +
> + COMMON /BLK1/ doub1, char1
> + COMMON /BLK2/ doub2, char2
> + DATA doub1, doub2 /1.111, 2.222/
> + DATA char1, char2 /'abcdef', 'ghijkl'/
> + END
> +
> +! SUBROUTINE
> + SUBROUTINE sub_block_data
> +
> + DOUBLE PRECISION doub1, doub2
> + CHARACTER*6 char1, char2
> +
> + COMMON /BLK1/ doub1, char1
> + COMMON /BLK2/ doub2, char2
> +
> + char1 = char2; ! BP_SUB
> + END
>
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: [PING] Re: [PATCH v3] Dwarf: Don't add nameless modules to partial symbol table.
2019-02-20 11:50 ` [PING] " Paweł Wódkowski
@ 2019-02-23 12:58 ` Joel Brobecker
0 siblings, 0 replies; 5+ messages in thread
From: Joel Brobecker @ 2019-02-23 12:58 UTC (permalink / raw)
To: Paweł Wódkowski; +Cc: gdb-patches
> Can someone take a look at this patch?
Sorry about the delay...
> > A name for BLOCK DATA in Fortran is optional. If no
> > name has been assigned, GDB will crash during read-in of DWARF
> > when BLOCK DATA is represented via DW_TAG_module.
> > BLOCK DATA is used for one-time initialization of
> > non-pointer variables in named common blocks.
> >
> > As of now there is no issue when gfortran is used as
> > DW_TAG_module is not emited. For Intel ifort the nameless
> > DW_TAG_module is present and have following form:
> >
> > ...
> > <1><dd>: Abbrev Number: 7 (DW_TAG_module)
> > <de> DW_AT_decl_line : 46
> > <df> DW_AT_decl_file : 1
> > <e0> DW_AT_description : (indirect string, offset: 0x110): block
> > data
> > <e4> DW_AT_high_pc : 0x402bb7
> > <ec> DW_AT_low_pc : 0x402bb7
> > ...
> >
> > Missing name lead to crash in add_partial_symbol function
> > during length calculation.
> >
> > 2016-06-15 Bernhard Heckel <bernhard.heckel@intel.com>
> >
> > gdb/Changelog:
> > * dwarf2read.c (add_partial_symbol): Skip nameless modules.
> >
> > gdb/testsuite/Changelog:
> > * gdb.fortran/block-data.f: New.
> > * gdb.fortran/block-data.exp: New.
> >
> > Changes from V1 to V2:
> > * No changes - just sent with other patches in series.
> >
> > Changes from V2 to V3:
> > * Decouple from other patches and send stand alone.
> > * Provide example DW_TAG_module in commit message.
> > * Fix mistakes in comments.
See comments below.
> > ---
> > gdb/dwarf2read.c | 13 +++++---
> > gdb/testsuite/gdb.fortran/block-data.exp | 49 ++++++++++++++++++++++++++++
> > gdb/testsuite/gdb.fortran/block-data.f | 56 ++++++++++++++++++++++++++++++++
> > 3 files changed, 113 insertions(+), 5 deletions(-)
> > create mode 100644 gdb/testsuite/gdb.fortran/block-data.exp
> > create mode 100644 gdb/testsuite/gdb.fortran/block-data.f
> >
> > diff --git a/gdb/dwarf2read.c b/gdb/dwarf2read.c
> > index d2a8cd44f9a5..389b21d27f3f 100644
> > --- a/gdb/dwarf2read.c
> > +++ b/gdb/dwarf2read.c
> > @@ -8995,11 +8995,14 @@ add_partial_symbol (struct partial_die_info *pdi, struct dwarf2_cu *cu)
> > 0, cu->language, objfile);
> > break;
> > case DW_TAG_module:
> > - add_psymbol_to_list (actual_name, strlen (actual_name),
> > - built_actual_name != NULL,
> > - MODULE_DOMAIN, LOC_TYPEDEF, -1,
> > - &objfile->global_psymbols,
> > - 0, cu->language, objfile);
> > + /* In Fortran 77 there might be a "BLOCK DATA" module available wihout
> > + any name. If so, we skip the module as it doesn't bring any value. */
> > + if (actual_name != nullptr)
> > + add_psymbol_to_list (actual_name, strlen (actual_name),
> > + built_actual_name != nullptr,
> > + MODULE_DOMAIN, LOC_TYPEDEF, -1,
> > + &objfile->global_psymbols,
> > + 0, cu->language, objfile);
I checked the DWARF 5 specifications, and it's clear from there
that not having a name is allowed. So this part of the change looks
OK to me.
I think we should also adjust read_module_type, however, to remove
a complaint:
module_name = dwarf2_name (die, cu);
if (!module_name)
complaint (_("DW_TAG_module has no name, offset %s"),
sect_offset_str (die->sect_off));
If it is allowed, then we shouldn't complain about it.
> > break;
> > case DW_TAG_class_type:
> > case DW_TAG_interface_type:
> > diff --git a/gdb/testsuite/gdb.fortran/block-data.exp b/gdb/testsuite/gdb.fortran/block-data.exp
> > new file mode 100644
> > index 000000000000..bb40be23d26b
> > --- /dev/null
> > +++ b/gdb/testsuite/gdb.fortran/block-data.exp
> > @@ -0,0 +1,49 @@
> > +# Copyright 2019 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/>.
> > +
> > +# This test is supposed to test anonymous block-data statement.
> > +
> > +if { [skip_fortran_tests] } { return -1 }
> > +
> > +standard_testfile .f
> > +load_lib "fortran.exp"
> > +
> > +if {[prepare_for_testing $testfile.exp $testfile $srcfile {debug f90}]} {
> > + return -1
> > +}
> > +
> > +if ![runto MAIN__] then {
> > + untested "couldn't run to breakpoint MAIN__"
> > + return -1
> > +}
> > +
> > +gdb_test "print doub1" "= 1.11\\d+" "print doub1, default values"
Can you escape the period in the floating point numbers above and below?
> > +gdb_test "print doub2" "= 2.22\\d+" "print doub2, default values"
> > +gdb_test "print char1" "= 'abcdef'" "print char1, default values"
> > +gdb_test "print char2" "= 'ghijkl'" "print char2, default values"
> > +
> > +gdb_breakpoint [gdb_get_line_number "! BP_BEFORE_SUB"]
> > +gdb_continue_to_breakpoint "! BP_BEFORE_SUB" ".*! BP_BEFORE_SUB.*"
> > +gdb_test "print doub1" "= 11.11\\d+" "print doub1, before sub"
> > +gdb_test "print doub2" "= 22.22\\d+" "print doub2, before sub"
> > +gdb_test "print char1" "= 'ABCDEF'" "print char1, before sub"
> > +gdb_test "print char2" "= 'GHIJKL'" "print char2, before sub"
> > +
> > +gdb_breakpoint [gdb_get_line_number "! BP_SUB"]
> > +gdb_continue_to_breakpoint "! BP_SUB" ".*! BP_SUB.*"
> > +gdb_test "print doub1" "= 11.11\\d+" "print char1, in sub"
> > +gdb_test "print doub2" "= 22.22\\d+" "print doub2, in sub"
> > +gdb_test "print char1" "= 'ABCDEF'" "print char1, in sub"
> > +gdb_test "print char2" "= 'GHIJKL'" "print char2, in sub"
--
Joel
^ permalink raw reply [flat|nested] 5+ messages in thread
* [PATCH v4] Dwarf: Don't add nameless modules to partial symbol table.
2019-01-05 11:44 [PATCH v3] Dwarf: Don't add nameless modules to partial symbol table Pawel Wodkowski
2019-02-20 11:50 ` [PING] " Paweł Wódkowski
@ 2019-02-27 14:18 ` Tomasz Kulasek
2019-06-11 19:24 ` Pedro Alves
1 sibling, 1 reply; 5+ messages in thread
From: Tomasz Kulasek @ 2019-02-27 14:18 UTC (permalink / raw)
To: gdb-patches; +Cc: Bernhard Heckel
From: Bernhard Heckel <bernhard.heckel@intel.com>
A name for BLOCK DATA in Fortran is optional. If no
name has been assigned, GDB will crash during read-in of DWARF
when BLOCK DATA is represented via DW_TAG_module.
BLOCK DATA is used for one-time initialization of
non-pointer variables in named common blocks.
As of now there is no issue when gfortran is used as
DW_TAG_module is not emited. For Intel ifort the nameless
DW_TAG_module is present and have following form:
...
<1><dd>: Abbrev Number: 7 (DW_TAG_module)
<de> DW_AT_decl_line : 46
<df> DW_AT_decl_file : 1
<e0> DW_AT_description : (indirect string, offset: 0x110): block
data
<e4> DW_AT_high_pc : 0x402bb7
<ec> DW_AT_low_pc : 0x402bb7
...
Missing name lead to crash in add_partial_symbol function
during length calculation.
2016-06-15 Bernhard Heckel <bernhard.heckel@intel.com>
gdb/Changelog:
* dwarf2read.c (add_partial_symbol): Skip nameless modules.
gdb/testsuite/Changelog:
* gdb.fortran/block-data.f: New.
* gdb.fortran/block-data.exp: New.
Changes from V1 to V2:
* No changes - just sent with other patches in series.
Changes from V2 to V3:
* Decouple from patch series.
* Provide example DW_TAG_module in commit message.
* Fix mistakes in comments.
Changes from V3 to V4:
* Dont complain if module have no name in read_module_type.
* Fix period excapeing in block-data.exp.
---
gdb/dwarf2read.c | 16 ++++-----
gdb/testsuite/gdb.fortran/block-data.exp | 49 ++++++++++++++++++++++++++++
gdb/testsuite/gdb.fortran/block-data.f | 56 ++++++++++++++++++++++++++++++++
3 files changed, 113 insertions(+), 8 deletions(-)
create mode 100644 gdb/testsuite/gdb.fortran/block-data.exp
create mode 100644 gdb/testsuite/gdb.fortran/block-data.f
diff --git a/gdb/dwarf2read.c b/gdb/dwarf2read.c
index 98f46e0..3b0274d 100644
--- a/gdb/dwarf2read.c
+++ b/gdb/dwarf2read.c
@@ -9018,11 +9018,14 @@ add_partial_symbol (struct partial_die_info *pdi, struct dwarf2_cu *cu)
0, cu->language, objfile);
break;
case DW_TAG_module:
- add_psymbol_to_list (actual_name, strlen (actual_name),
- built_actual_name != NULL,
- MODULE_DOMAIN, LOC_TYPEDEF, -1,
- psymbol_placement::GLOBAL,
- 0, cu->language, objfile);
+ /* In Fortran 77 there might be a "BLOCK DATA" module available wihout
+ any name. If so, we skip the module as it doesn't bring any value. */
+ if (actual_name != nullptr)
+ add_psymbol_to_list (actual_name, strlen (actual_name),
+ built_actual_name != NULL,
+ MODULE_DOMAIN, LOC_TYPEDEF, -1,
+ psymbol_placement::GLOBAL,
+ 0, cu->language, objfile);
break;
case DW_TAG_class_type:
case DW_TAG_interface_type:
@@ -16929,9 +16932,6 @@ read_module_type (struct die_info *die, struct dwarf2_cu *cu)
struct type *type;
module_name = dwarf2_name (die, cu);
- if (!module_name)
- complaint (_("DW_TAG_module has no name, offset %s"),
- sect_offset_str (die->sect_off));
type = init_type (objfile, TYPE_CODE_MODULE, 0, module_name);
return set_die_type (die, type, cu);
diff --git a/gdb/testsuite/gdb.fortran/block-data.exp b/gdb/testsuite/gdb.fortran/block-data.exp
new file mode 100644
index 0000000..389ae15
--- /dev/null
+++ b/gdb/testsuite/gdb.fortran/block-data.exp
@@ -0,0 +1,49 @@
+# Copyright 2019 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/>.
+
+# This test is supposed to test anonymous block-data statement.
+
+if { [skip_fortran_tests] } { return -1 }
+
+standard_testfile .f
+load_lib "fortran.exp"
+
+if {[prepare_for_testing $testfile.exp $testfile $srcfile {debug f90}]} {
+ return -1
+}
+
+if ![runto MAIN__] then {
+ untested "couldn't run to breakpoint MAIN__"
+ return -1
+}
+
+gdb_test "print doub1" "= 1\.11\\d+" "print doub1, default values"
+gdb_test "print doub2" "= 2\.22\\d+" "print doub2, default values"
+gdb_test "print char1" "= 'abcdef'" "print char1, default values"
+gdb_test "print char2" "= 'ghijkl'" "print char2, default values"
+
+gdb_breakpoint [gdb_get_line_number "! BP_BEFORE_SUB"]
+gdb_continue_to_breakpoint "! BP_BEFORE_SUB" ".*! BP_BEFORE_SUB.*"
+gdb_test "print doub1" "= 11\.11\\d+" "print doub1, before sub"
+gdb_test "print doub2" "= 22\.22\\d+" "print doub2, before sub"
+gdb_test "print char1" "= 'ABCDEF'" "print char1, before sub"
+gdb_test "print char2" "= 'GHIJKL'" "print char2, before sub"
+
+gdb_breakpoint [gdb_get_line_number "! BP_SUB"]
+gdb_continue_to_breakpoint "! BP_SUB" ".*! BP_SUB.*"
+gdb_test "print doub1" "= 11\.11\\d+" "print char1, in sub"
+gdb_test "print doub2" "= 22\.22\\d+" "print doub2, in sub"
+gdb_test "print char1" "= 'ABCDEF'" "print char1, in sub"
+gdb_test "print char2" "= 'GHIJKL'" "print char2, in sub"
diff --git a/gdb/testsuite/gdb.fortran/block-data.f b/gdb/testsuite/gdb.fortran/block-data.f
new file mode 100644
index 0000000..5f50e64
--- /dev/null
+++ b/gdb/testsuite/gdb.fortran/block-data.f
@@ -0,0 +1,56 @@
+! Copyright 2019 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/>.
+!
+! Test if GDB can handle block data without global name
+!
+! MAIN
+ PROGRAM bdata
+ DOUBLE PRECISION doub1, doub2
+ CHARACTER*6 char1, char2
+
+ COMMON /BLK1/ doub1, char1
+ COMMON /BLK2/ doub2, char2
+
+ doub1 = 11.111
+ doub2 = 22.222
+ char1 = 'ABCDEF'
+ char2 = 'GHIJKL'
+ CALL sub_block_data ! BP_BEFORE_SUB
+ STOP
+ END
+
+! BLOCK DATA
+ BLOCK DATA
+
+ DOUBLE PRECISION doub1, doub2
+ CHARACTER*6 char1, char2
+
+ COMMON /BLK1/ doub1, char1
+ COMMON /BLK2/ doub2, char2
+ DATA doub1, doub2 /1.111, 2.222/
+ DATA char1, char2 /'abcdef', 'ghijkl'/
+ END
+
+! SUBROUTINE
+ SUBROUTINE sub_block_data
+
+ DOUBLE PRECISION doub1, doub2
+ CHARACTER*6 char1, char2
+
+ COMMON /BLK1/ doub1, char1
+ COMMON /BLK2/ doub2, char2
+
+ char1 = char2; ! BP_SUB
+ END
--
2.7.4
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: [PATCH v4] Dwarf: Don't add nameless modules to partial symbol table.
2019-02-27 14:18 ` [PATCH v4] " Tomasz Kulasek
@ 2019-06-11 19:24 ` Pedro Alves
0 siblings, 0 replies; 5+ messages in thread
From: Pedro Alves @ 2019-06-11 19:24 UTC (permalink / raw)
To: Tomasz Kulasek, gdb-patches; +Cc: Bernhard Heckel, Joel Brobecker
Hi,
Looks like this has been reviewed a number of times over
the years but never managed to get merged. I was asked
internally at Red Hat to take a look, and I've merged it now.
I confirmed that this version addresses Joel's comments to v3.
See more below.
On 2/27/19 2:21 PM, Tomasz Kulasek wrote:
> From: Bernhard Heckel <bernhard.heckel@intel.com>
>
> A name for BLOCK DATA in Fortran is optional. If no
> name has been assigned, GDB will crash during read-in of DWARF
> when BLOCK DATA is represented via DW_TAG_module.
> BLOCK DATA is used for one-time initialization of
> non-pointer variables in named common blocks.
>
> As of now there is no issue when gfortran is used as
> DW_TAG_module is not emited. For Intel ifort the nameless
> DW_TAG_module is present and have following form:
>
> ...
> <1><dd>: Abbrev Number: 7 (DW_TAG_module)
> <de> DW_AT_decl_line : 46
> <df> DW_AT_decl_file : 1
> <e0> DW_AT_description : (indirect string, offset: 0x110): block
> data
> <e4> DW_AT_high_pc : 0x402bb7
> <ec> DW_AT_low_pc : 0x402bb7
> ...
>
> Missing name lead to crash in add_partial_symbol function
> during length calculation.
>
> 2016-06-15 Bernhard Heckel <bernhard.heckel@intel.com>
>
> gdb/Changelog:
> * dwarf2read.c (add_partial_symbol): Skip nameless modules.
>
> gdb/testsuite/Changelog:
> * gdb.fortran/block-data.f: New.
> * gdb.fortran/block-data.exp: New.
>
> Changes from V1 to V2:
> * No changes - just sent with other patches in series.
>
> Changes from V2 to V3:
> * Decouple from patch series.
> * Provide example DW_TAG_module in commit message.
> * Fix mistakes in comments.
>
> Changes from V3 to V4:
> * Dont complain if module have no name in read_module_type.
> * Fix period excapeing in block-data.exp.
> ---
> gdb/dwarf2read.c | 16 ++++-----
> gdb/testsuite/gdb.fortran/block-data.exp | 49 ++++++++++++++++++++++++++++
> gdb/testsuite/gdb.fortran/block-data.f | 56 ++++++++++++++++++++++++++++++++
> 3 files changed, 113 insertions(+), 8 deletions(-)
> create mode 100644 gdb/testsuite/gdb.fortran/block-data.exp
> create mode 100644 gdb/testsuite/gdb.fortran/block-data.f
>
> diff --git a/gdb/dwarf2read.c b/gdb/dwarf2read.c
> index 98f46e0..3b0274d 100644
> --- a/gdb/dwarf2read.c
> +++ b/gdb/dwarf2read.c
> @@ -9018,11 +9018,14 @@ add_partial_symbol (struct partial_die_info *pdi, struct dwarf2_cu *cu)
> 0, cu->language, objfile);
> break;
> case DW_TAG_module:
> - add_psymbol_to_list (actual_name, strlen (actual_name),
> - built_actual_name != NULL,
> - MODULE_DOMAIN, LOC_TYPEDEF, -1,
> - psymbol_placement::GLOBAL,
> - 0, cu->language, objfile);
> + /* In Fortran 77 there might be a "BLOCK DATA" module available wihout
Typo "wihout".
> + any name. If so, we skip the module as it doesn't bring any value. */
> + if (actual_name != nullptr)
> + add_psymbol_to_list (actual_name, strlen (actual_name),
> + built_actual_name != NULL,
> + MODULE_DOMAIN, LOC_TYPEDEF, -1,
> + psymbol_placement::GLOBAL,
> + 0, cu->language, objfile);
> break;
> case DW_TAG_class_type:
> case DW_TAG_interface_type:
> @@ -16929,9 +16932,6 @@ read_module_type (struct die_info *die, struct dwarf2_cu *cu)
> struct type *type;
>
> module_name = dwarf2_name (die, cu);
> - if (!module_name)
> - complaint (_("DW_TAG_module has no name, offset %s"),
> - sect_offset_str (die->sect_off));
> type = init_type (objfile, TYPE_CODE_MODULE, 0, module_name);
>
> return set_die_type (die, type, cu);
> diff --git a/gdb/testsuite/gdb.fortran/block-data.exp b/gdb/testsuite/gdb.fortran/block-data.exp
> new file mode 100644
> index 0000000..389ae15
> --- /dev/null
> +++ b/gdb/testsuite/gdb.fortran/block-data.exp
> @@ -0,0 +1,49 @@
> +# Copyright 2019 Free Software Foundation, Inc.
This file was first posted in 2016, so the copyright years should include
that.
> +
> +gdb_test "print doub1" "= 1\.11\\d+" "print doub1, default values"
> +gdb_test "print doub2" "= 2\.22\\d+" "print doub2, default values"
> +gdb_test "print char1" "= 'abcdef'" "print char1, default values"
> +gdb_test "print char2" "= 'ghijkl'" "print char2, default values"
> +
> +gdb_breakpoint [gdb_get_line_number "! BP_BEFORE_SUB"]
> +gdb_continue_to_breakpoint "! BP_BEFORE_SUB" ".*! BP_BEFORE_SUB.*"
> +gdb_test "print doub1" "= 11\.11\\d+" "print doub1, before sub"
> +gdb_test "print doub2" "= 22\.22\\d+" "print doub2, before sub"
> +gdb_test "print char1" "= 'ABCDEF'" "print char1, before sub"
> +gdb_test "print char2" "= 'GHIJKL'" "print char2, before sub"
> +
> +gdb_breakpoint [gdb_get_line_number "! BP_SUB"]
> +gdb_continue_to_breakpoint "! BP_SUB" ".*! BP_SUB.*"
> +gdb_test "print doub1" "= 11\.11\\d+" "print char1, in sub"
^^^^^
I checked for duplicate test names, and that revealed the typo
above. Should be "doub1" instead of "char1".
But better is to use with_test_prefix instead.
> +gdb_test "print doub2" "= 22\.22\\d+" "print doub2, in sub"
> +gdb_test "print char1" "= 'ABCDEF'" "print char1, in sub"
> +gdb_test "print char2" "= 'GHIJKL'" "print char2, in sub"
> +!
> +! Test if GDB can handle block data without global name
Missing period.
I've now applied the patch, with the issues above fixed, along
with a few other editorial tweaks to commit log and comments.
Note that I've extended the intro comment in the .exp file.
Here's what I merged.
From a5fd13a91534b1c79a4b61995894a5bb4f08d3b0 Mon Sep 17 00:00:00 2001
From: Bernhard Heckel <bernhard.heckel@intel.com>
Date: Wed, 27 Feb 2019 15:21:12 +0100
Subject: [PATCH] Dwarf: Don't add nameless modules to partial symbol table
A name for BLOCK DATA in Fortran is optional. If no name has been
assigned, GDB crashes during read-in of DWARF when BLOCK DATA is
represented via DW_TAG_module. BLOCK DATA is used for one-time
initialization of non-pointer variables in named common blocks.
As of now there is no issue when gfortran is used as DW_TAG_module is
not emitted. However, with Intel ifort the nameless DW_TAG_module is
present and has the following form:
...
<1><dd>: Abbrev Number: 7 (DW_TAG_module)
<de> DW_AT_decl_line : 46
<df> DW_AT_decl_file : 1
<e0> DW_AT_description : (indirect string, offset: 0x110): block
data
<e4> DW_AT_high_pc : 0x402bb7
<ec> DW_AT_low_pc : 0x402bb7
...
The missing name leads to a crash in add_partial_symbol, during length
calculation.
gdb/ChangeLog:
2019-06-11 Bernhard Heckel <bernhard.heckel@intel.com>
* dwarf2read.c (add_partial_symbol): Skip nameless modules.
gdb/testsuite/Changelog:
2019-06-11 Bernhard Heckel <bernhard.heckel@intel.com>
* gdb.fortran/block-data.f: New.
* gdb.fortran/block-data.exp: New.
---
gdb/dwarf2read.c | 17 +++++----
gdb/testsuite/gdb.fortran/block-data.exp | 63 ++++++++++++++++++++++++++++++++
gdb/testsuite/gdb.fortran/block-data.f | 56 ++++++++++++++++++++++++++++
3 files changed, 128 insertions(+), 8 deletions(-)
create mode 100644 gdb/testsuite/gdb.fortran/block-data.exp
create mode 100644 gdb/testsuite/gdb.fortran/block-data.f
diff --git a/gdb/dwarf2read.c b/gdb/dwarf2read.c
index 06c319dac02..4cf9fcfa218 100644
--- a/gdb/dwarf2read.c
+++ b/gdb/dwarf2read.c
@@ -9018,11 +9018,15 @@ add_partial_symbol (struct partial_die_info *pdi, struct dwarf2_cu *cu)
0, cu->language, objfile);
break;
case DW_TAG_module:
- add_psymbol_to_list (actual_name, strlen (actual_name),
- built_actual_name != NULL,
- MODULE_DOMAIN, LOC_TYPEDEF, -1,
- psymbol_placement::GLOBAL,
- 0, cu->language, objfile);
+ /* With Fortran 77 there might be a "BLOCK DATA" module
+ available without any name. If so, we skip the module as it
+ doesn't bring any value. */
+ if (actual_name != nullptr)
+ add_psymbol_to_list (actual_name, strlen (actual_name),
+ built_actual_name != NULL,
+ MODULE_DOMAIN, LOC_TYPEDEF, -1,
+ psymbol_placement::GLOBAL,
+ 0, cu->language, objfile);
break;
case DW_TAG_class_type:
case DW_TAG_interface_type:
@@ -16955,9 +16959,6 @@ read_module_type (struct die_info *die, struct dwarf2_cu *cu)
struct type *type;
module_name = dwarf2_name (die, cu);
- if (!module_name)
- complaint (_("DW_TAG_module has no name, offset %s"),
- sect_offset_str (die->sect_off));
type = init_type (objfile, TYPE_CODE_MODULE, 0, module_name);
return set_die_type (die, type, cu);
diff --git a/gdb/testsuite/gdb.fortran/block-data.exp b/gdb/testsuite/gdb.fortran/block-data.exp
new file mode 100644
index 00000000000..8da8c8824c8
--- /dev/null
+++ b/gdb/testsuite/gdb.fortran/block-data.exp
@@ -0,0 +1,63 @@
+# Copyright 2016-2019 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/>.
+
+# Test anonymous block-data statements.
+
+# A name for BLOCK DATA in Fortran is optional. BLOCK DATA is used
+# for one-time initialization of non-pointer variables in named common
+# blocks. GDB used to crash with 'Intel ifort'-generated code, which
+# outputs nameless DW_TAG_module, unlike with gfortran which just
+# doesn't emit DW_TAG_module in this case.
+
+if { [skip_fortran_tests] } { return -1 }
+
+standard_testfile .f
+load_lib "fortran.exp"
+
+if {[prepare_for_testing $testfile.exp $testfile $srcfile {debug f90}]} {
+ return -1
+}
+
+if ![runto MAIN__] then {
+ untested "couldn't run to breakpoint MAIN__"
+ return -1
+}
+
+with_test_prefix "default values" {
+ gdb_test "print doub1" "= 1\.11\\d+"
+ gdb_test "print doub2" "= 2\.22\\d+"
+ gdb_test "print char1" "= 'abcdef'"
+ gdb_test "print char2" "= 'ghijkl'"
+}
+
+gdb_breakpoint [gdb_get_line_number "! BP_BEFORE_SUB"]
+gdb_continue_to_breakpoint "! BP_BEFORE_SUB" ".*! BP_BEFORE_SUB.*"
+
+with_test_prefix "before sub" {
+ gdb_test "print doub1" "= 11\.11\\d+"
+ gdb_test "print doub2" "= 22\.22\\d+"
+ gdb_test "print char1" "= 'ABCDEF'"
+ gdb_test "print char2" "= 'GHIJKL'"
+}
+
+gdb_breakpoint [gdb_get_line_number "! BP_SUB"]
+gdb_continue_to_breakpoint "! BP_SUB" ".*! BP_SUB.*"
+
+with_test_prefix "in sub" {
+ gdb_test "print doub1" "= 11\.11\\d+"
+ gdb_test "print doub2" "= 22\.22\\d+"
+ gdb_test "print char1" "= 'ABCDEF'"
+ gdb_test "print char2" "= 'GHIJKL'"
+}
diff --git a/gdb/testsuite/gdb.fortran/block-data.f b/gdb/testsuite/gdb.fortran/block-data.f
new file mode 100644
index 00000000000..acb2c5a8591
--- /dev/null
+++ b/gdb/testsuite/gdb.fortran/block-data.f
@@ -0,0 +1,56 @@
+! Copyright 2016-2019 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/>.
+!
+! Check that GDB can handle block data with no global name.
+!
+! MAIN
+ PROGRAM bdata
+ DOUBLE PRECISION doub1, doub2
+ CHARACTER*6 char1, char2
+
+ COMMON /BLK1/ doub1, char1
+ COMMON /BLK2/ doub2, char2
+
+ doub1 = 11.111
+ doub2 = 22.222
+ char1 = 'ABCDEF'
+ char2 = 'GHIJKL'
+ CALL sub_block_data ! BP_BEFORE_SUB
+ STOP
+ END
+
+! BLOCK DATA
+ BLOCK DATA
+
+ DOUBLE PRECISION doub1, doub2
+ CHARACTER*6 char1, char2
+
+ COMMON /BLK1/ doub1, char1
+ COMMON /BLK2/ doub2, char2
+ DATA doub1, doub2 /1.111, 2.222/
+ DATA char1, char2 /'abcdef', 'ghijkl'/
+ END
+
+! SUBROUTINE
+ SUBROUTINE sub_block_data
+
+ DOUBLE PRECISION doub1, doub2
+ CHARACTER*6 char1, char2
+
+ COMMON /BLK1/ doub1, char1
+ COMMON /BLK2/ doub2, char2
+
+ char1 = char2; ! BP_SUB
+ END
--
2.14.5
^ permalink raw reply [flat|nested] 5+ messages in thread
end of thread, other threads:[~2019-06-11 19:24 UTC | newest]
Thread overview: 5+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2019-01-05 11:44 [PATCH v3] Dwarf: Don't add nameless modules to partial symbol table Pawel Wodkowski
2019-02-20 11:50 ` [PING] " Paweł Wódkowski
2019-02-23 12:58 ` Joel Brobecker
2019-02-27 14:18 ` [PATCH v4] " Tomasz Kulasek
2019-06-11 19:24 ` Pedro Alves
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).