* [PATCH] testsuite: sigall.exp: handle SIGLOST/SIGPWR conflict in sparc64 targets
@ 2013-10-23 14:04 Jose E. Marchesi
2013-12-03 11:58 ` Jose E. Marchesi
0 siblings, 1 reply; 4+ messages in thread
From: Jose E. Marchesi @ 2013-10-23 14:04 UTC (permalink / raw)
To: gdb-patches
Hi.
In sparc64 SIGLOST and SIGPWR have the same signal number. The
following patch fixes gdb.base/sigall.exp accordingly.
Tested in sparc64-unknown-linux-gnu.
2013-10-23 Jose E. Marchesi <jose.marchesi@oracle.com>
* gdb.base/sigall.exp (test_one_sig): gdb identifies SIGLOST as a
SIGPWR in sparc64.
* gdb.base/sigall.c (main): In some targets SIGLOST and SIGPWR
have the same signal number. Handle this situation.
diff --git a/gdb/testsuite/gdb.base/sigall.c b/gdb/testsuite/gdb.base/sigall.c
index 110029d..68f3ecb 100644
--- a/gdb/testsuite/gdb.base/sigall.c
+++ b/gdb/testsuite/gdb.base/sigall.c
@@ -1083,7 +1083,7 @@ return 0;
int
gen_PWR ()
{
-#ifdef SIGPWR
+#if defined(SIGPWR) && (!defined (SIGLOST) || SIGPWR != SIGLOST)
kill (getpid (), SIGPWR);
#else
handle_PWR (0);
@@ -1662,7 +1662,7 @@ main ()
#ifdef SIGUSR2
signal (SIGUSR2, handle_USR2);
#endif
-#ifdef SIGPWR
+#if defined(SIGPWR) && (!defined (SIGLOST) || SIGPWR != SIGLOST)
signal (SIGPWR, handle_PWR);
#endif
#if defined (SIGPOLL) && (!defined (SIGIO) || SIGPOLL != SIGIO)
diff --git a/gdb/testsuite/gdb.base/sigall.exp b/gdb/testsuite/gdb.base/sigall.exp
index b1b8597..4bfb199 100644
--- a/gdb/testsuite/gdb.base/sigall.exp
+++ b/gdb/testsuite/gdb.base/sigall.exp
@@ -39,12 +39,20 @@ proc test_one_sig {nextsig} {
set need_another_continue 1
set missed_handler 0
if $this_sig_supported then {
+ set esig $thissig
+
if { $thissig == "IO" } {
setup_xfail "i*86-pc-linuxoldld-gnu" "i*86-pc-linuxaout-gnu"
}
+ # In SPARC64 systems SIGLOST==SIGPWR and gdb identifies the raised
+ # signal as PWR.
+ if { $thissig == "LOST" && [ istarget "sparc64-*-linux*" ] } then {
+ set esig "PWR"
+ }
+
gdb_test "continue" \
- "Continuing.*Program received signal SIG$thissig.*" \
- "get signal $thissig"
+ "Continuing.*Program received signal SIG$esig.*" \
+ "get signal $esig"
}
if [ istarget "alpha-dec-osf3*" ] then {
# OSF/1-3.x is unable to continue with a job control stop signal.
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: [PATCH] testsuite: sigall.exp: handle SIGLOST/SIGPWR conflict in sparc64 targets
2013-10-23 14:04 [PATCH] testsuite: sigall.exp: handle SIGLOST/SIGPWR conflict in sparc64 targets Jose E. Marchesi
@ 2013-12-03 11:58 ` Jose E. Marchesi
2013-12-03 12:07 ` Mark Kettenis
0 siblings, 1 reply; 4+ messages in thread
From: Jose E. Marchesi @ 2013-12-03 11:58 UTC (permalink / raw)
To: gdb-patches
ping
Hi.
In sparc64 SIGLOST and SIGPWR have the same signal number. The
following patch fixes gdb.base/sigall.exp accordingly.
Tested in sparc64-unknown-linux-gnu.
2013-10-23 Jose E. Marchesi <jose.marchesi@oracle.com>
* gdb.base/sigall.exp (test_one_sig): gdb identifies SIGLOST as a
SIGPWR in sparc64.
* gdb.base/sigall.c (main): In some targets SIGLOST and SIGPWR
have the same signal number. Handle this situation.
diff --git a/gdb/testsuite/gdb.base/sigall.c b/gdb/testsuite/gdb.base/sigall.c
index 110029d..68f3ecb 100644
--- a/gdb/testsuite/gdb.base/sigall.c
+++ b/gdb/testsuite/gdb.base/sigall.c
@@ -1083,7 +1083,7 @@ return 0;
int
gen_PWR ()
{
-#ifdef SIGPWR
+#if defined(SIGPWR) && (!defined (SIGLOST) || SIGPWR != SIGLOST)
kill (getpid (), SIGPWR);
#else
handle_PWR (0);
@@ -1662,7 +1662,7 @@ main ()
#ifdef SIGUSR2
signal (SIGUSR2, handle_USR2);
#endif
-#ifdef SIGPWR
+#if defined(SIGPWR) && (!defined (SIGLOST) || SIGPWR != SIGLOST)
signal (SIGPWR, handle_PWR);
#endif
#if defined (SIGPOLL) && (!defined (SIGIO) || SIGPOLL != SIGIO)
diff --git a/gdb/testsuite/gdb.base/sigall.exp b/gdb/testsuite/gdb.base/sigall.exp
index b1b8597..4bfb199 100644
--- a/gdb/testsuite/gdb.base/sigall.exp
+++ b/gdb/testsuite/gdb.base/sigall.exp
@@ -39,12 +39,20 @@ proc test_one_sig {nextsig} {
set need_another_continue 1
set missed_handler 0
if $this_sig_supported then {
+ set esig $thissig
+
if { $thissig == "IO" } {
setup_xfail "i*86-pc-linuxoldld-gnu" "i*86-pc-linuxaout-gnu"
}
+ # In SPARC64 systems SIGLOST==SIGPWR and gdb identifies the raised
+ # signal as PWR.
+ if { $thissig == "LOST" && [ istarget "sparc64-*-linux*" ] } then {
+ set esig "PWR"
+ }
+
gdb_test "continue" \
- "Continuing.*Program received signal SIG$thissig.*" \
- "get signal $thissig"
+ "Continuing.*Program received signal SIG$esig.*" \
+ "get signal $esig"
}
if [ istarget "alpha-dec-osf3*" ] then {
# OSF/1-3.x is unable to continue with a job control stop signal.
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: [PATCH] testsuite: sigall.exp: handle SIGLOST/SIGPWR conflict in sparc64 targets
2013-12-03 11:58 ` Jose E. Marchesi
@ 2013-12-03 12:07 ` Mark Kettenis
2013-12-03 12:53 ` Jose E. Marchesi
0 siblings, 1 reply; 4+ messages in thread
From: Mark Kettenis @ 2013-12-03 12:07 UTC (permalink / raw)
To: jose.marchesi; +Cc: gdb-patches
> From: jose.marchesi@oracle.com (Jose E. Marchesi)
> Date: Tue, 03 Dec 2013 12:59:02 +0100
>
> ping
>
> Hi.
>
> In sparc64 SIGLOST and SIGPWR have the same signal number. The
> following patch fixes gdb.base/sigall.exp accordingly.
>
> Tested in sparc64-unknown-linux-gnu.
>
> 2013-10-23 Jose E. Marchesi <jose.marchesi@oracle.com>
>
> * gdb.base/sigall.exp (test_one_sig): gdb identifies SIGLOST as a
> SIGPWR in sparc64.
>
> * gdb.base/sigall.c (main): In some targets SIGLOST and SIGPWR
> have the same signal number. Handle this situation.
Small nit:
> diff --git a/gdb/testsuite/gdb.base/sigall.exp b/gdb/testsuite/gdb.base/sigall.exp
> index b1b8597..4bfb199 100644
> --- a/gdb/testsuite/gdb.base/sigall.exp
> +++ b/gdb/testsuite/gdb.base/sigall.exp
> @@ -39,12 +39,20 @@ proc test_one_sig {nextsig} {
> set need_another_continue 1
> set missed_handler 0
> if $this_sig_supported then {
> + set esig $thissig
> +
> if { $thissig == "IO" } {
> setup_xfail "i*86-pc-linuxoldld-gnu" "i*86-pc-linuxaout-gnu"
> }
> + # In SPARC64 systems SIGLOST==SIGPWR and gdb identifies the raised
> + # signal as PWR.
Change that into "On Linux SPARC64 systems..."
With that change, this is fine with me.
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: [PATCH] testsuite: sigall.exp: handle SIGLOST/SIGPWR conflict in sparc64 targets
2013-12-03 12:07 ` Mark Kettenis
@ 2013-12-03 12:53 ` Jose E. Marchesi
0 siblings, 0 replies; 4+ messages in thread
From: Jose E. Marchesi @ 2013-12-03 12:53 UTC (permalink / raw)
To: Mark Kettenis; +Cc: gdb-patches
> + # In SPARC64 systems SIGLOST==SIGPWR and gdb identifies the raised
> + # signal as PWR.
Change that into "On Linux SPARC64 systems..."
With that change, this is fine with me.
Ok. Patch amended below and committed.
(By mistake I pushed a wrong date in the testsuite/ChangeLog entry and
had to fix that with a new commit. Sorry about that!.)
2013-12-03 Jose E. Marchesi <jose.marchesi@oracle.com>
* gdb.base/sigall.exp (test_one_sig): gdb identifies SIGLOST as a
SIGPWR in sparc64.
* gdb.base/sigall.c (main): In some targets SIGLOST and SIGPWR
have the same signal number. Handle this situation.
diff --git a/gdb/testsuite/gdb.base/sigall.c b/gdb/testsuite/gdb.base/sigall.c
index 110029d..68f3ecb 100644
--- a/gdb/testsuite/gdb.base/sigall.c
+++ b/gdb/testsuite/gdb.base/sigall.c
@@ -1083,7 +1083,7 @@ return 0;
int
gen_PWR ()
{
-#ifdef SIGPWR
+#if defined(SIGPWR) && (!defined (SIGLOST) || SIGPWR != SIGLOST)
kill (getpid (), SIGPWR);
#else
handle_PWR (0);
@@ -1662,7 +1662,7 @@ main ()
#ifdef SIGUSR2
signal (SIGUSR2, handle_USR2);
#endif
-#ifdef SIGPWR
+#if defined(SIGPWR) && (!defined (SIGLOST) || SIGPWR != SIGLOST)
signal (SIGPWR, handle_PWR);
#endif
#if defined (SIGPOLL) && (!defined (SIGIO) || SIGPOLL != SIGIO)
diff --git a/gdb/testsuite/gdb.base/sigall.exp b/gdb/testsuite/gdb.base/sigall.exp
index b1b8597..d531479 100644
--- a/gdb/testsuite/gdb.base/sigall.exp
+++ b/gdb/testsuite/gdb.base/sigall.exp
@@ -39,12 +39,20 @@ proc test_one_sig {nextsig} {
set need_another_continue 1
set missed_handler 0
if $this_sig_supported then {
+ set esig $thissig
+
if { $thissig == "IO" } {
setup_xfail "i*86-pc-linuxoldld-gnu" "i*86-pc-linuxaout-gnu"
}
+ # On Linux SPARC64 systems SIGLOST==SIGPWR and gdb identifies
+ # the raised signal as PWR.
+ if { $thissig == "LOST" && [ istarget "sparc64-*-linux*" ] } then {
+ set esig "PWR"
+ }
+
gdb_test "continue" \
- "Continuing.*Program received signal SIG$thissig.*" \
- "get signal $thissig"
+ "Continuing.*Program received signal SIG$esig.*" \
+ "get signal $esig"
}
if [ istarget "alpha-dec-osf3*" ] then {
# OSF/1-3.x is unable to continue with a job control stop signal.
^ permalink raw reply [flat|nested] 4+ messages in thread
end of thread, other threads:[~2013-12-03 12:53 UTC | newest]
Thread overview: 4+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2013-10-23 14:04 [PATCH] testsuite: sigall.exp: handle SIGLOST/SIGPWR conflict in sparc64 targets Jose E. Marchesi
2013-12-03 11:58 ` Jose E. Marchesi
2013-12-03 12:07 ` Mark Kettenis
2013-12-03 12:53 ` Jose E. Marchesi
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox