Skip to contentSkip to conversation

ldd01: using 'ppc' to check powerpc arch

PATCH In review Han Pingtian 3 participants 6 messages Sep 11, 2014 – Sep 13, 2014

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 ldd01: using 'ppc' to check powerpc arch

Hello,
The ldd01 always failed on some powerpc systems of ours. I have figured out a patch. Please take a look. Thanks in advance.

On some powerpc system, the 'uname -a' doesn't show 'powerpc' but will show 'ppc'. So we need using it to check the powerpc arch.

---
 testcases/commands/ade/ldd/ldd01 | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
testcases/commands/ade/ldd/ldd01 +1 -1
index c60d3ac..569902e 100755
@@ -47,7 +47,7 @@ do_setup()
 	mkdir $TCtmp
 
 	# Check for ppc64 architecture or x86_64 architecture
-	if uname -a | grep -iq powerpc; then
+	if uname -a | grep -iq 'ppc\|powerpc'; then
 		file lddfile1.obj | grep 64-bit >/dev/null 2>&1
 		if [ $? -eq 0 ]; then
 			CFLAGS="-m64"
Signature · 13 lines
1.9.3


------------------------------------------------------------------------------
Want excitement?
Manually upgrade your production database.
When you want reliability, choose Perforce
Perforce version control. Predictably reliable.
http://pubads.g.doubleclick.net/gampad/clk?id=157508191&iu=/4140/ostg.clktrk
_______________________________________________
Ltp-list mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/ltp-list
#1 lore raw

Han Pingtian <[email protected]> wrote on 09/11/2014 07:57:09 AM:

Quoted 43 lines

Han Pingtian <[email protected]>
09/11/2014 07:57 AM

To

[email protected]

cc

Subject

[LTP] [PATCH] ldd01: using 'ppc' to check powerpc arch

Hello,
The ldd01 always failed on some powerpc systems of ours. I have figured out a patch. Please take a look. Thanks in advance.

On some powerpc system, the 'uname -a' doesn't show 'powerpc' but will show 'ppc'. So we need using it to check the powerpc arch.

---
 testcases/commands/ade/ldd/ldd01 | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
testcases/commands/ renamed from testcases/commands/ade/ldd/ldd01 +0 -0

ade/ldd/ldd01
index c60d3ac..569902e 100755

testcases/commands/ade/ldd/ldd01 +1 -1
@@ -47,7 +47,7 @@ do_setup()
    mkdir $TCtmp
 
    # Check for ppc64 architecture or x86_64 architecture
-   if uname -a | grep -iq powerpc; then
+   if uname -a | grep -iq 'ppc\|powerpc'; then
       file lddfile1.obj | grep 64-bit >/dev/null 2>&1
       if [ $? -eq 0 ]; then
          CFLAGS="-m64"
Signature · 1 line
1.9.3

Verified on PPC machine this fix helps

Quoted 2 lines

------------------------------------------------------------------------------

Quoted 5 lines

Want excitement?
Manually upgrade your production database.
When you want reliability, choose Perforce
Perforce version control. Predictably reliable.

http://pubads.g.doubleclick.net/gampad/clk?id=157508191&iu=/4140/ostg.clktrk

Quoted 5 lines

_______________________________________________
Ltp-list mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/ltp-list

------------------------------------------------------------------------------
Want excitement?
Manually upgrade your production database.
When you want reliability, choose Perforce
Perforce version control. Predictably reliable.
http://pubads.g.doubleclick.net/gampad/clk?id=157508191&iu=/4140/ostg.clktrk

_______________________________________________
Ltp-list mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/ltp-list

#2 lore raw
In reply to Han Pingtian
Quoted 18 lines · Han Pingtian

Hello,
The ldd01 always failed on some powerpc systems of ours. I have figured out a patch. Please take a look. Thanks in advance.

On some powerpc system, the 'uname -a' doesn't show 'powerpc' but will show 'ppc'. So we need using it to check the powerpc arch.

---
 testcases/commands/ade/ldd/ldd01 | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
testcases/commands/ade/ldd/ldd01 +0 -0
index c60d3ac..569902e 100755

testcases/commands/ade/ldd/ldd01 do_setup() · line 53

Quoted 8 lines
testcases/commands/ade/ldd/ldd01 +1 -1
@@ -47,7 +47,7 @@ do_setup()
 	mkdir $TCtmp
 
 	# Check for ppc64 architecture or x86_64 architecture
-	if uname -a | grep -iq powerpc; then
+	if uname -a | grep -iq 'ppc\|powerpc'; then
 		file lddfile1.obj | grep 64-bit >/dev/null 2>&1
 		if [ $? -eq 0 ]; then
 			CFLAGS="-m64"

Hi,

what is the purpose of those -m32/-m64 flags in this test? Can't we just compile it with LTP makefiles using whatever CFLAGS were passed to configure?

Would you be willing to try attached patch, which moves the build of .so and a.out into Makefile? I have tested it on x86 with and without -m32 and it worked fine. Or alternatively, can you describe the ppc system where this is failing, so I could try to reproduce it myself?

Regards,
Jan

From 1ea966268e7a55f21d9254b62462d4adb40c7a0a Mon Sep 17 00:00:00 2001
Message-Id: <1ea966268e7a55f21d9254b62462d4adb40c7a0a.1410445852.git.jstancek@redhat.com>
From: Jan Stancek <[email protected]>
Date: Thu, 11 Sep 2014 14:57:37 +0200
Subject: [PATCH] v1

---
 testcases/commands/ade/ldd/Makefile   |   13 +++++++++++--
 testcases/commands/ade/ldd/ldd01      |   28 ++++------------------------
 testcases/commands/ade/ldd/lddfile1.c |    2 +-
 testcases/commands/ade/ldd/lddfile2.c |    2 +-
 testcases/commands/ade/ldd/lddfile3.c |    2 +-
 testcases/commands/ade/ldd/lddfile4.c |    2 +-
 testcases/commands/ade/ldd/lddfile5.c |    2 +-
 7 files changed, 20 insertions(+), 31 deletions(-)
testcases/commands/ade/ldd/Makefile +11 -2
index eae06cd..7119290 100644
@@ -28,9 +28,18 @@ CPPFLAGS		+= -fpic
 
 INSTALL_TARGETS		:= ldd01
 
-MAKE_TARGETS		:= $(patsubst $(abs_srcdir)/%.c,%.obj,$(wildcard $(abs_srcdir)/*.c))
+LDD_C_FILES		:= $(wildcard $(abs_srcdir)/lddfile*.c)
+LDD_SO_FILES		:= $(patsubst $(abs_srcdir)/%.c,%.obj.so,$(LDD_C_FILES))
+MAKE_TARGETS		:= $(LDD_SO_FILES) lddfile.out
+CLEAN_TARGETS		+= *.obj
 
-$(MAKE_TARGETS): %.obj: %.o
+%.obj: %.o
 	mv $^ $@
 
+%.obj.so: %.obj
+	$(CC) $(CFLAGS) -shared -o $@ $^
+
+lddfile.out: main.obj $(LDD_SO_FILES)
+	$(CC) $(CFLAGS) -O -o $@ $(LDD_SO_FILES) main.obj
+
 include $(top_srcdir)/include/mk/generic_leaf_target.mk
testcases/commands/ade/ldd/ldd01 +4 -24
index c60d3ac..54188f7 100755
@@ -35,42 +35,22 @@ $trace_logic
 CC=${CC:=gcc}
 LDD=${LDD:=ldd}
 TCdat=${TCdat:-`pwd`}
-TCtmp=${TCtmp:-/tmp/ldd01-$$}
 
 do_cleanup()
 {
-	rm -rf $TCtmp
+	:
 }
 
 do_setup()
 {
-	mkdir $TCtmp
-
-	# Check for ppc64 architecture or x86_64 architecture
-	if uname -a | grep -iq powerpc; then
-		file lddfile1.obj | grep 64-bit >/dev/null 2>&1
-		if [ $? -eq 0 ]; then
-			CFLAGS="-m64"
-		fi
-	elif uname -a | grep -iq x86_64; then
-		file lddfile1.obj | grep 32-bit >/dev/null 2>&1
-		if [ $? -eq 0 ]; then
-			CFLAGS="-m32"
-		fi
-	fi
-	$CC $CFLAGS -shared -o $TCtmp/lddfile1.obj.so $TCdat/lddfile1.obj
-	$CC $CFLAGS -shared -o $TCtmp/lddfile2.obj.so $TCdat/lddfile2.obj
-	$CC $CFLAGS -shared -o $TCtmp/lddfile3.obj.so $TCdat/lddfile3.obj
-	$CC $CFLAGS -shared -o $TCtmp/lddfile4.obj.so $TCdat/lddfile4.obj
-	$CC $CFLAGS -shared -o $TCtmp/lddfile5.obj.so $TCdat/lddfile5.obj
-	$CC $CFLAGS -O -o $TCtmp/a.out $TCtmp/lddfile*.obj.so $TCdat/main.obj
+	:
 }
 
 do_test()
 {
 	echo "  ASSERTION 1  "
 
-	$LDD $TCtmp/a.out | grep -E "lddfile1.obj.so|lddfile2.obj.so|lddfile3.obj.so|lddfile4.obj.so|lddfile5.obj.so"
+	$LDD lddfile.out | grep -E "lddfile1.obj.so|lddfile2.obj.so|lddfile3.obj.so|lddfile4.obj.so|lddfile5.obj.so"
 	if [ $? -eq 0 ]; then
 		echo "ASSERTION #1 PASS"
 	else
@@ -81,7 +61,7 @@ do_test()
 
 	echo " ASSERTION 2  "
 
-	$LDD -v $TCtmp/a.out | grep -E "GLIBC|lddfile1.obj.so|lddfile2.obj.so|lddfile3.obj.so|lddfile4.obj.so|lddfile5.obj.so"
+	$LDD -v lddfile.out | grep -E "GLIBC|lddfile1.obj.so|lddfile2.obj.so|lddfile3.obj.so|lddfile4.obj.so|lddfile5.obj.so"
 	if [ $? -eq 0 ]; then
 		echo "ASSERTION #2 PASS"
 	else
testcases/commands/ade/ldd/lddfile1.c +1 -1
index 377f7a3..ccb5786 100644
@@ -1,5 +1,5 @@
 #include <stdio.h>
-void file1()
+void file1(void)
 {
 	printf("Control in function %s\n", __func__);
 }
testcases/commands/ade/ldd/lddfile2.c +1 -1
index 5fd2b4b..20aede3 100644
@@ -1,5 +1,5 @@
 #include <stdio.h>
-void file2()
+void file2(void)
 {
 	printf("Control in function %s\n", __func__);
 }
testcases/commands/ade/ldd/lddfile3.c +1 -1
index d6b4cc9..a4529f7 100644
@@ -1,5 +1,5 @@
 #include <stdio.h>
-void file3()
+void file3(void)
 {
 	printf("Control in function %s\n", __func__);
 }
testcases/commands/ade/ldd/lddfile4.c +1 -1
index 1db89bb..9dc339d 100644
@@ -1,5 +1,5 @@
 #include <stdio.h>
-void file4()
+void file4(void)
 {
 	printf("Control in function %s\n", __func__);
 }
testcases/commands/ade/ldd/lddfile5.c +1 -1
index 2f46cdb..3fbb7f6 100644
@@ -1,5 +1,5 @@
 #include <stdio.h>
-void file5()
+void file5(void)
 {
 	printf("Control in function %s\n", __func__);
 }
Signature · 1 line
1.7.1

------------------------------------------------------------------------------
Want excitement?
Manually upgrade your production database.
When you want reliability, choose Perforce
Perforce version control. Predictably reliable.
http://pubads.g.doubleclick.net/gampad/clk?id=157508191&iu=/4140/ostg.clktrk

_______________________________________________
Ltp-list mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/ltp-list

#3 lore raw
Quoted 2 lines · Jan Stancek

Would you be willing to try attached patch, which moves the build of .so and a.out into Makefile?

I have tested your patch but ldd cannot find the lib files:

% sudo ./ldd01
ASSERTION 1
lddfile1.obj.so => not found lddfile2.obj.so => not found lddfile3.obj.so => not found lddfile4.obj.so => not found lddfile5.obj.so => not found
ASSERTION #1 PASS
ASSERTION 2
lddfile1.obj.so => not found lddfile2.obj.so => not found lddfile3.obj.so => not found lddfile4.obj.so => not found lddfile5.obj.so => not found
libc.so.6 (GLIBC_2.3) => /lib64/libc.so.6 libc.so.6 (GLIBC_2.4) => /lib64/libc.so.6 ld64.so.1 (GLIBC_PRIVATE) => /lib64/ld64.so.1 ld64.so.1 (GLIBC_2.3) => /lib64/ld64.so.1
ASSERTION #2 PASS
TEST PASSES

on the contray, with my patch lib files can be found:

% ./ldd01
ASSERTION 1

        /tmp/ldd01-9497/lddfile1.obj.so (0x00000fffad800000)
        /tmp/ldd01-9497/lddfile2.obj.so (0x00000fffad7e0000)
        /tmp/ldd01-9497/lddfile3.obj.so (0x00000fffad7c0000)
        /tmp/ldd01-9497/lddfile4.obj.so (0x00000fffad7a0000)
        /tmp/ldd01-9497/lddfile5.obj.so (0x00000fffad780000)
ASSERTION #1 PASS
 ASSERTION 2  
        /tmp/ldd01-9497/lddfile1.obj.so (0x00000fffb1d20000)
        /tmp/ldd01-9497/lddfile2.obj.so (0x00000fffb1d00000)
        /tmp/ldd01-9497/lddfile3.obj.so (0x00000fffb1ce0000)
        /tmp/ldd01-9497/lddfile4.obj.so (0x00000fffb1cc0000)
        /tmp/ldd01-9497/lddfile5.obj.so (0x00000fffb1ca0000)
                libc.so.6 (GLIBC_2.3) => /lib64/libc.so.6
                libc.so.6 (GLIBC_2.4) => /lib64/libc.so.6
        /tmp/ldd01-9497/lddfile1.obj.so:
                libc.so.6 (GLIBC_2.3) => /lib64/libc.so.6
                libc.so.6 (GLIBC_2.4) => /lib64/libc.so.6
        /tmp/ldd01-9497/lddfile2.obj.so:
                libc.so.6 (GLIBC_2.3) => /lib64/libc.so.6
                libc.so.6 (GLIBC_2.4) => /lib64/libc.so.6
        /tmp/ldd01-9497/lddfile3.obj.so:
                libc.so.6 (GLIBC_2.3) => /lib64/libc.so.6
                libc.so.6 (GLIBC_2.4) => /lib64/libc.so.6
        /tmp/ldd01-9497/lddfile4.obj.so:
                libc.so.6 (GLIBC_2.3) => /lib64/libc.so.6
                libc.so.6 (GLIBC_2.4) => /lib64/libc.so.6
        /tmp/ldd01-9497/lddfile5.obj.so:
                libc.so.6 (GLIBC_2.3) => /lib64/libc.so.6
                libc.so.6 (GLIBC_2.4) => /lib64/libc.so.6
                ld64.so.1 (GLIBC_PRIVATE) => /lib64/ld64.so.1
                ld64.so.1 (GLIBC_2.3) => /lib64/ld64.so.1

ASSERTION #2 PASS
TEST PASSES

Quoted 3 lines

Or alternatively, can you describe
the ppc system where this is failing, so I could try to reproduce it myself?

This is a power 6 system, which running a beta version rhel5.11. Without my patch, it failed with those messages:

% ./ldd01
/usr/bin/ld: warning: powerpc:common64 architecture of input file /testcases/ltp/testcases/commands/ade/ldd/lddfile1.obj' is incompatible with powerpc:common output /usr/bin/ld: warning: powerpc:common64 architecture of input file /testcases/ltp/testcases/commands/ade/ldd/lddfile2.obj' is incompatible with powerpc:common output /usr/bin/ld: warning: powerpc:common64 architecture of input file /testcases/ltp/testcases/commands/ade/ldd/lddfile3.obj' is incompatible with powerpc:common output /usr/bin/ld: warning: powerpc:common64 architecture of input file /testcases/ltp/testcases/commands/ade/ldd/lddfile4.obj' is incompatible with powerpc:common output /usr/bin/ld: warning: powerpc:common64 architecture of input file /testcases/ltp/testcases/commands/ade/ldd/lddfile5.obj' is incompatible with powerpc:common output /usr/bin/ld: warning: powerpc:common64 architecture of input file /testcases/ltp/testcases/commands/ade/ldd/main.obj' is incompatible with powerpc:common output /testcases/ltp/testcases/commands/ade/ldd/main.obj: In function `main':
/home/hpt/ltp-full-20130503.64b/testcases/commands/ade/ldd/main.c:11: relocation truncated to fit: R_PPC64_REL24 against file1' /home/hpt/ltp-full-20130503.64b/testcases/commands/ade/ldd/main.c:12: relocation truncated to fit: R_PPC64_REL24 against file2' /home/hpt/ltp-full-20130503.64b/testcases/commands/ade/ldd/main.c:13: relocation truncated to fit: R_PPC64_REL24 against file3' /home/hpt/ltp-full-20130503.64b/testcases/commands/ade/ldd/main.c:14: relocation truncated to fit: R_PPC64_REL24 against file4' /home/hpt/ltp-full-20130503.64b/testcases/commands/ade/ldd/main.c:15: relocation truncated to fit: R_PPC64_REL24 against file5' /home/hpt/ltp-full-20130503.64b/testcases/commands/ade/ldd/main.c:16: relocation truncated to fit: R_PPC64_REL24 against puts'
collect2: ld returned 1 exit status
ASSERTION 1
ldd: /tmp/ldd01-14133/a.out: No such file or directory
ASSERTION #1 FAIL

Quoted 158 lines
Quoted 1 line

From 1ea966268e7a55f21d9254b62462d4adb40c7a0a Mon Sep 17 00:00:00 2001

Message-Id: <1ea966268e7a55f21d9254b62462d4adb40c7a0a.1410445852.git.jstancek@redhat.com>
From: Jan Stancek <[email protected]>
Date: Thu, 11 Sep 2014 14:57:37 +0200
Subject: [PATCH] v1

---
 testcases/commands/ade/ldd/Makefile   |   13 +++++++++++--
 testcases/commands/ade/ldd/ldd01      |   28 ++++------------------------
 testcases/commands/ade/ldd/lddfile1.c |    2 +-
 testcases/commands/ade/ldd/lddfile2.c |    2 +-
 testcases/commands/ade/ldd/lddfile3.c |    2 +-
 testcases/commands/ade/ldd/lddfile4.c |    2 +-
 testcases/commands/ade/ldd/lddfile5.c |    2 +-
 7 files changed, 20 insertions(+), 31 deletions(-)
testcases/commands/ade/ldd/Makefile +11 -2
index eae06cd..7119290 100644
@@ -28,9 +28,18 @@ CPPFLAGS		+= -fpic
 
 INSTALL_TARGETS		:= ldd01
 
-MAKE_TARGETS		:= $(patsubst $(abs_srcdir)/%.c,%.obj,$(wildcard $(abs_srcdir)/*.c))
+LDD_C_FILES		:= $(wildcard $(abs_srcdir)/lddfile*.c)
+LDD_SO_FILES		:= $(patsubst $(abs_srcdir)/%.c,%.obj.so,$(LDD_C_FILES))
+MAKE_TARGETS		:= $(LDD_SO_FILES) lddfile.out
+CLEAN_TARGETS		+= *.obj
 
-$(MAKE_TARGETS): %.obj: %.o
+%.obj: %.o
 	mv $^ $@
 
+%.obj.so: %.obj
+	$(CC) $(CFLAGS) -shared -o $@ $^
+
+lddfile.out: main.obj $(LDD_SO_FILES)
+	$(CC) $(CFLAGS) -O -o $@ $(LDD_SO_FILES) main.obj
+
 include $(top_srcdir)/include/mk/generic_leaf_target.mk
testcases/commands/ade/ldd/ldd01 +4 -24
index c60d3ac..54188f7 100755
@@ -35,42 +35,22 @@ $trace_logic
 CC=${CC:=gcc}
 LDD=${LDD:=ldd}
 TCdat=${TCdat:-`pwd`}
-TCtmp=${TCtmp:-/tmp/ldd01-$$}
 
 do_cleanup()
 {
-	rm -rf $TCtmp
+	:
 }
 
 do_setup()
 {
-	mkdir $TCtmp
-
-	# Check for ppc64 architecture or x86_64 architecture
-	if uname -a | grep -iq powerpc; then
-		file lddfile1.obj | grep 64-bit >/dev/null 2>&1
-		if [ $? -eq 0 ]; then
-			CFLAGS="-m64"
-		fi
-	elif uname -a | grep -iq x86_64; then
-		file lddfile1.obj | grep 32-bit >/dev/null 2>&1
-		if [ $? -eq 0 ]; then
-			CFLAGS="-m32"
-		fi
-	fi
-	$CC $CFLAGS -shared -o $TCtmp/lddfile1.obj.so $TCdat/lddfile1.obj
-	$CC $CFLAGS -shared -o $TCtmp/lddfile2.obj.so $TCdat/lddfile2.obj
-	$CC $CFLAGS -shared -o $TCtmp/lddfile3.obj.so $TCdat/lddfile3.obj
-	$CC $CFLAGS -shared -o $TCtmp/lddfile4.obj.so $TCdat/lddfile4.obj
-	$CC $CFLAGS -shared -o $TCtmp/lddfile5.obj.so $TCdat/lddfile5.obj
-	$CC $CFLAGS -O -o $TCtmp/a.out $TCtmp/lddfile*.obj.so $TCdat/main.obj
+	:
 }
 
 do_test()
 {
 	echo "  ASSERTION 1  "
 
-	$LDD $TCtmp/a.out | grep -E "lddfile1.obj.so|lddfile2.obj.so|lddfile3.obj.so|lddfile4.obj.so|lddfile5.obj.so"
+	$LDD lddfile.out | grep -E "lddfile1.obj.so|lddfile2.obj.so|lddfile3.obj.so|lddfile4.obj.so|lddfile5.obj.so"
 	if [ $? -eq 0 ]; then
 		echo "ASSERTION #1 PASS"
 	else
@@ -81,7 +61,7 @@ do_test()
 
 	echo " ASSERTION 2  "
 
-	$LDD -v $TCtmp/a.out | grep -E "GLIBC|lddfile1.obj.so|lddfile2.obj.so|lddfile3.obj.so|lddfile4.obj.so|lddfile5.obj.so"
+	$LDD -v lddfile.out | grep -E "GLIBC|lddfile1.obj.so|lddfile2.obj.so|lddfile3.obj.so|lddfile4.obj.so|lddfile5.obj.so"
 	if [ $? -eq 0 ]; then
 		echo "ASSERTION #2 PASS"
 	else
testcases/commands/ade/ldd/lddfile1.c +1 -1
index 377f7a3..ccb5786 100644
@@ -1,5 +1,5 @@
 #include <stdio.h>
-void file1()
+void file1(void)
 {
 	printf("Control in function %s\n", __func__);
 }
testcases/commands/ade/ldd/lddfile2.c +1 -1
index 5fd2b4b..20aede3 100644
@@ -1,5 +1,5 @@
 #include <stdio.h>
-void file2()
+void file2(void)
 {
 	printf("Control in function %s\n", __func__);
 }
testcases/commands/ade/ldd/lddfile3.c +1 -1
index d6b4cc9..a4529f7 100644
@@ -1,5 +1,5 @@
 #include <stdio.h>
-void file3()
+void file3(void)
 {
 	printf("Control in function %s\n", __func__);
 }
testcases/commands/ade/ldd/lddfile4.c +1 -1
index 1db89bb..9dc339d 100644
@@ -1,5 +1,5 @@
 #include <stdio.h>
-void file4()
+void file4(void)
 {
 	printf("Control in function %s\n", __func__);
 }
testcases/commands/ade/ldd/lddfile5.c +1 -1
index 2f46cdb..3fbb7f6 100644
@@ -1,5 +1,5 @@
 #include <stdio.h>
-void file5()
+void file5(void)
 {
 	printf("Control in function %s\n", __func__);
 }
Signature · 1 line
1.7.1

------------------------------------------------------------------------------
Want excitement?
Manually upgrade your production database.
When you want reliability, choose Perforce
Perforce version control. Predictably reliable.
http://pubads.g.doubleclick.net/gampad/clk?id=157508191&iu=/4140/ostg.clktrk
_______________________________________________
Ltp-list mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/ltp-list

#4 lore raw

----- Original Message -----

Quoted 10 lines

From: "Han Pingtian" <[email protected]>
To: [email protected]
Sent: Friday, 12 September, 2014 9:28:45 AM
Subject: Re: [LTP] [PATCH] ldd01: using 'ppc' to check powerpc arch

Quoted 2 lines · Jan Stancek

Would you be willing to try attached patch, which moves the build of .so and a.out into Makefile?

I have tested your patch but ldd cannot find the lib files:

That's only matter of setting correct LD_LIBRARY_PATH:

$ env LD_LIBRARY_PATH=.:$LD_LIBRARY_PATH ./ldd01 
  ASSERTION 1  
	lddfile1.obj.so => ./lddfile1.obj.so (0x00007f06504db000)
	lddfile2.obj.so => ./lddfile2.obj.so (0x00007f06502d9000)
	lddfile3.obj.so => ./lddfile3.obj.so (0x00007f06500d8000)
	lddfile4.obj.so => ./lddfile4.obj.so (0x00007f064fed7000)
	lddfile5.obj.so => ./lddfile5.obj.so (0x00007f064fcd5000)
ASSERTION #1 PASS
 ASSERTION 2  
	lddfile1.obj.so => ./lddfile1.obj.so (0x00007ffe913e9000)
	lddfile2.obj.so => ./lddfile2.obj.so (0x00007ffe911e7000)
	lddfile3.obj.so => ./lddfile3.obj.so (0x00007ffe90fe6000)
	lddfile4.obj.so => ./lddfile4.obj.so (0x00007ffe90de5000)
	lddfile5.obj.so => ./lddfile5.obj.so (0x00007ffe90be3000)
		libc.so.6 (GLIBC_2.2.5) => /lib64/libc.so.6
		libc.so.6 (GLIBC_2.3.4) => /lib64/libc.so.6
	./lddfile1.obj.so:
		libc.so.6 (GLIBC_2.2.5) => /lib64/libc.so.6
		libc.so.6 (GLIBC_2.3.4) => /lib64/libc.so.6
	./lddfile2.obj.so:
		libc.so.6 (GLIBC_2.2.5) => /lib64/libc.so.6
		libc.so.6 (GLIBC_2.3.4) => /lib64/libc.so.6
	./lddfile3.obj.so:
		libc.so.6 (GLIBC_2.2.5) => /lib64/libc.so.6
		libc.so.6 (GLIBC_2.3.4) => /lib64/libc.so.6
	./lddfile4.obj.so:
		libc.so.6 (GLIBC_2.2.5) => /lib64/libc.so.6
		libc.so.6 (GLIBC_2.3.4) => /lib64/libc.so.6
	./lddfile5.obj.so:
		libc.so.6 (GLIBC_2.2.5) => /lib64/libc.so.6
		libc.so.6 (GLIBC_2.3.4) => /lib64/libc.so.6
		ld-linux-x86-64.so.2 (GLIBC_PRIVATE) => /lib64/ld-linux-x86-64.so.2
		ld-linux-x86-64.so.2 (GLIBC_2.3) => /lib64/ld-linux-x86-64.so.2
ASSERTION #2 PASS
TEST PASSES
Quoted 6 lines
Quoted 4 lines

Or alternatively, can you describe
the ppc system where this is failing, so I could try to reproduce it myself?

This is a power 6 system, which running a beta version rhel5.11. Without my patch, it failed with those messages:

It also fails (with or without your patch) when built as 32-bit on RHEL7 ppc64:

# ./ldd01 
/usr/bin/ld: powerpc:common architecture of input file `/root/ltp/testcases/commands/ade/ldd/lddfile1.obj' is incompatible with powerpc:common64 output
/root/ltp/testcases/commands/ade/ldd/lddfile1.obj: In function `file1':
/root/ltp/testcases/commands/ade/ldd/lddfile1.c:3: undefined reference to `_GLOBAL_OFFSET_TABLE_'
/root/ltp/testcases/commands/ade/ldd/lddfile1.c:3: undefined reference to `_GLOBAL_OFFSET_TABLE_'
collect2: error: ld returned 1 exit status
/usr/bin/ld: powerpc:common architecture of input file `/root/ltp/testcases/commands/ade/ldd/lddfile2.obj' is incompatible with powerpc:common64 output
/root/ltp/testcases/commands/ade/ldd/lddfile2.obj: In function `file2':
/root/ltp/testcases/commands/ade/ldd/lddfile2.c:3: undefined reference to `_GLOBAL_OFFSET_TABLE_'
/root/ltp/testcases/commands/ade/ldd/lddfile2.c:3: undefined reference to `_GLOBAL_OFFSET_TABLE_'
collect2: error: ld returned 1 exit status
/usr/bin/ld: powerpc:common architecture of input file `/root/ltp/testcases/commands/ade/ldd/lddfile3.obj' is incompatible with powerpc:common64 output
/root/ltp/testcases/commands/ade/ldd/lddfile3.obj: In function `file3':
/root/ltp/testcases/commands/ade/ldd/lddfile3.c:3: undefined reference to `_GLOBAL_OFFSET_TABLE_'
/root/ltp/testcases/commands/ade/ldd/lddfile3.c:3: undefined reference to `_GLOBAL_OFFSET_TABLE_'
collect2: error: ld returned 1 exit status
/usr/bin/ld: powerpc:common architecture of input file `/root/ltp/testcases/commands/ade/ldd/lddfile4.obj' is incompatible with powerpc:common64 output
/root/ltp/testcases/commands/ade/ldd/lddfile4.obj: In function `file4':
/root/ltp/testcases/commands/ade/ldd/lddfile4.c:3: undefined reference to `_GLOBAL_OFFSET_TABLE_'
/root/ltp/testcases/commands/ade/ldd/lddfile4.c:3: undefined reference to `_GLOBAL_OFFSET_TABLE_'
collect2: error: ld returned 1 exit status
/usr/bin/ld: powerpc:common architecture of input file `/root/ltp/testcases/commands/ade/ldd/lddfile5.obj' is incompatible with powerpc:common64 output
/root/ltp/testcases/commands/ade/ldd/lddfile5.obj: In function `file5':
/root/ltp/testcases/commands/ade/ldd/lddfile5.c:3: undefined reference to `_GLOBAL_OFFSET_TABLE_'
/root/ltp/testcases/commands/ade/ldd/lddfile5.c:3: undefined reference to `_GLOBAL_OFFSET_TABLE_'
collect2: error: ld returned 1 exit status
gcc: error: /tmp/ldd01-26565/lddfile*.obj.so: No such file or directory
  ASSERTION 1  
ldd: /tmp/ldd01-26565/a.out: No such file or directory
ASSERTION #1 FAIL

which is why I'd prefer to avoid all that -m32/-m64 logic in ldd01.

Regards,
Jan

------------------------------------------------------------------------------
Want excitement?
Manually upgrade your production database. When you want reliability, choose Perforce Perforce version control. Predictably reliable.
http://pubads.g.doubleclick.net/gampad/clk?id=157508191&iu=/4140/ostg.clktrk
_______________________________________________
Ltp-list mailing list [email protected]
https://lists.sourceforge.net/lists/listinfo/ltp-list

#5 lore raw
Quoted 10 lines · Jan Stancek

----- Original Message -----

Quoted 6 lines

From: "Han Pingtian" <[email protected]>
To: [email protected]
Sent: Friday, 12 September, 2014 9:28:45 AM
Subject: Re: [LTP] [PATCH] ldd01: using 'ppc' to check powerpc arch

On Thu, Sep 11, 2014 at 04:41:29PM +0200, Jan Stancek wrote:

Quoted 8 lines
Quoted 4 lines
Quoted 2 lines

Would you be willing to try attached patch, which moves the build of .so and a.out into Makefile?

I have tested your patch but ldd cannot find the lib files:

That's only matter of setting correct LD_LIBRARY_PATH:

$ env LD_LIBRARY_PATH=.:$LD_LIBRARY_PATH ./ldd01 

Yes, it works after setting LD_LIBRARY_PATH like this on our systems.

Quoted 9 lines
Quoted 6 lines
Quoted 4 lines

Or alternatively, can you describe
the ppc system where this is failing, so I could try to reproduce it myself?

This is a power 6 system, which running a beta version rhel5.11. Without my patch, it failed with those messages:

It also fails (with or without your patch) when built as 32-bit on RHEL7 ppc64:

I have tested your patch with 32bit on the system, it works fine.

------------------------------------------------------------------------------
Want excitement?
Manually upgrade your production database.
When you want reliability, choose Perforce
Perforce version control. Predictably reliable.
http://pubads.g.doubleclick.net/gampad/clk?id=157508191&iu=/4140/ostg.clktrk
_______________________________________________
Ltp-list mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/ltp-list