Mirror of the gdb-patches mailing list
 help / color / mirror / Atom feed
* [committed][gdb/testsuite] Fix exec_is_pie with gold linker
@ 2020-05-25 10:02 Tom de Vries
  2020-05-25 14:54 ` Pedro Alves
  0 siblings, 1 reply; 3+ messages in thread
From: Tom de Vries @ 2020-05-25 10:02 UTC (permalink / raw)
  To: gdb-patches

Hi,

When running test-case gdb.base/break-interp.exp with target board gold, we
run into:
...
gdb compile failed, pie failed to generate PIE executable
...

The problem is that the proc exec_is_pie uses the PIE flag in the readelf -d
output, which doesn't seem to be set by the gold linker.

Instead, use the "Type" field in the readelf -h output.

Tested on x86_64-linux.

Committed to trunk.

Thanks,
- Tom

[gdb/testsuite] Fix exec_is_pie with gold linker

gdb/testsuite/ChangeLog:

2020-05-25  Tom de Vries  <tdevries@suse.de>

	PR testsuite/26031
	* lib/gdb.exp (exec_is_pie): Test readelf -h output.

---
 gdb/testsuite/lib/gdb.exp | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/gdb/testsuite/lib/gdb.exp b/gdb/testsuite/lib/gdb.exp
index f7d20bd94f..7177be941b 100644
--- a/gdb/testsuite/lib/gdb.exp
+++ b/gdb/testsuite/lib/gdb.exp
@@ -5509,11 +5509,12 @@ proc exec_is_pie { executable } {
 	return -1
     }
     set readelf_program [gdb_find_readelf]
-    set res [catch {exec $readelf_program -d $executable} output]
+    set res [catch {exec $readelf_program -h $executable} output]
     if { $res != 0 } {
 	return -1
     }
-    set res [regexp -line {\(FLAGS_1\).*Flags:.* PIE($| )} $output]
+    set res [regexp -line {^[ \t]*Type:[ \t]*DYN \(Shared object file\)$} \
+		 $output]
     if { $res == 1 } {
 	return 1
     }


^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: [committed][gdb/testsuite] Fix exec_is_pie with gold linker
  2020-05-25 10:02 [committed][gdb/testsuite] Fix exec_is_pie with gold linker Tom de Vries
@ 2020-05-25 14:54 ` Pedro Alves
  2020-05-25 15:29   ` Tom de Vries
  0 siblings, 1 reply; 3+ messages in thread
From: Pedro Alves @ 2020-05-25 14:54 UTC (permalink / raw)
  To: Tom de Vries, gdb-patches

On 5/25/20 11:02 AM, Tom de Vries wrote:
> Hi,
> 
> When running test-case gdb.base/break-interp.exp with target board gold, we
> run into:
> ...
> gdb compile failed, pie failed to generate PIE executable
> ...
> 
> The problem is that the proc exec_is_pie uses the PIE flag in the readelf -d
> output, which doesn't seem to be set by the gold linker.

Seems like a Gold bug?  A comment in exec_is_pie explaining why we don't
do the more obvious thing would be helpful, IMO.

> @@ -5509,11 +5509,12 @@ proc exec_is_pie { executable } {
>  	return -1
>      }
>      set readelf_program [gdb_find_readelf]
> -    set res [catch {exec $readelf_program -d $executable} output]
> +    set res [catch {exec $readelf_program -h $executable} output]
>      if { $res != 0 } {
>  	return -1
>      }
> -    set res [regexp -line {\(FLAGS_1\).*Flags:.* PIE($| )} $output]
> +    set res [regexp -line {^[ \t]*Type:[ \t]*DYN \(Shared object file\)$} \
> +		 $output]
>      if { $res == 1 } {
>  	return 1
>      }
> 

Thanks,
Pedro Alves



^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: [committed][gdb/testsuite] Fix exec_is_pie with gold linker
  2020-05-25 14:54 ` Pedro Alves
@ 2020-05-25 15:29   ` Tom de Vries
  0 siblings, 0 replies; 3+ messages in thread
From: Tom de Vries @ 2020-05-25 15:29 UTC (permalink / raw)
  To: Pedro Alves, gdb-patches

[-- Attachment #1: Type: text/plain, Size: 755 bytes --]

On 25-05-2020 16:54, Pedro Alves wrote:
> On 5/25/20 11:02 AM, Tom de Vries wrote:
>> Hi,
>>
>> When running test-case gdb.base/break-interp.exp with target board gold, we
>> run into:
>> ...
>> gdb compile failed, pie failed to generate PIE executable
>> ...
>>
>> The problem is that the proc exec_is_pie uses the PIE flag in the readelf -d
>> output, which doesn't seem to be set by the gold linker.
> 
> Seems like a Gold bug?

Yeah, Andreas Schwab had the same comment in bugzilla, so I filed
PR26039 - "gold doesn't add PIE flag" (
https://sourceware.org/bugzilla/show_bug.cgi?id=26039 ).	

> A comment in exec_is_pie explaining why we don't
> do the more obvious thing would be helpful, IMO.
> 

Fixed in attached patch, committed.

Thanks,
- Tom


[-- Attachment #2: 0001-gdb-testsuite-Add-comment-in-exec_is_pie.patch --]
[-- Type: text/x-patch, Size: 856 bytes --]

[gdb/testsuite] Add comment in exec_is_pie

Add comment to exec_is_pie explaining why readelf -d output is not used.

gdb/testsuite/ChangeLog:

2020-05-25  Tom de Vries  <tdevries@suse.de>

	* lib/gdb.exp (exec_is_pie): Add comment.

---
 gdb/testsuite/lib/gdb.exp | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/gdb/testsuite/lib/gdb.exp b/gdb/testsuite/lib/gdb.exp
index 7177be941b..8e22941f0b 100644
--- a/gdb/testsuite/lib/gdb.exp
+++ b/gdb/testsuite/lib/gdb.exp
@@ -5509,6 +5509,9 @@ proc exec_is_pie { executable } {
 	return -1
     }
     set readelf_program [gdb_find_readelf]
+    # We're not testing readelf -d | grep "FLAGS_1.*Flags:.*PIE"
+    # because the PIE flag is not set by all versions of gold, see PR
+    # binutils/26039.
     set res [catch {exec $readelf_program -h $executable} output]
     if { $res != 0 } {
 	return -1

^ permalink raw reply	[flat|nested] 3+ messages in thread

end of thread, other threads:[~2020-05-25 15:29 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2020-05-25 10:02 [committed][gdb/testsuite] Fix exec_is_pie with gold linker Tom de Vries
2020-05-25 14:54 ` Pedro Alves
2020-05-25 15:29   ` Tom de Vries

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox