[PATCH 2/2] powerpc: unrel_branch_check.sh: enable the use of llvm-objdump v9, 10 or 11

Stephen Rothwell sfr at canb.auug.org.au
Wed Aug 12 18:10:36 AEST 2020


Currently, using llvm-objtool, this script just silently succeeds without
actually do the intended checking.  So this updates it to work properly.

Firstly, llvm-objdump does not add target symbol names to the end
of branches in its asm output, so we have to drop the branch to
__start_initialization_multiplatform using its address.

Secondly, v9 and 10 specify branch targets as .+<offset>, so we convert
those to actual addresses.

Thirdly, v10 and 11 error out on a vmlinux if given the -R option
complaining that it is "not a dynamic object".  The -R does not make
any difference to the asm output, so remove it.

Lastly, v11 produces asm that is very similar to Gnu objtool (at least
as far as branches are concerned), so no further changes are necessary
to make it work.

Cc: Nicholas Piggin <npiggin at gmail.com>
Cc: Bill Wendling <morbo at google.com>
Signed-off-by: Stephen Rothwell <sfr at canb.auug.org.au>
---
 arch/powerpc/tools/unrel_branch_check.sh | 34 ++++++++++++++++++++----
 1 file changed, 29 insertions(+), 5 deletions(-)

diff --git a/arch/powerpc/tools/unrel_branch_check.sh b/arch/powerpc/tools/unrel_branch_check.sh
index 0369eb2e7e4b..8301efee1e6c 100755
--- a/arch/powerpc/tools/unrel_branch_check.sh
+++ b/arch/powerpc/tools/unrel_branch_check.sh
@@ -18,12 +18,16 @@ if [ "$end_intr" = "0x" ]; then
 	exit 0
 fi
 
-$objdump -R -D --no-show-raw-insn --start-address="$kstart" --stop-address="$end_intr" "$vmlinux" |
+# we know that there is a correct branch to
+# __start_initialization_multiplatform, so find its address
+# so we can exclude it.
+sim=0x$($nm -p "$vmlinux" |
+	sed -E -n '/\s+[[:alpha:]]\s+__start_initialization_multiplatform\s*$/{s///p;q}')
+
+$objdump -D --no-show-raw-insn --start-address="$kstart" --stop-address="$end_intr" "$vmlinux" |
 sed -E -n '
 # match lines that start with a kernel address
 /^c[0-9a-f]*:\s*b/ {
-	# drop a target that we do not care about
-	/\<__start_initialization_multiplatform>/d
 	# drop branches via ctr or lr
 	/\<b.?.?(ct|l)r/d
 	# cope with some differences between Clang and GNU objdumps
@@ -33,14 +37,34 @@ sed -E -n '
 	s/\s0x/ /
 	s/://
 	# format for the loop below
-	s/^(\S+)\s+(\S+)\s+(\S+)\s*(\S*).*$/\1:\2:0x\3:\4/
+	s/^(\S+)\s+(\S+)\s+(\S+)\s*(\S*).*$/\1:\2:\3:\4/
 	# strip out condition registers
-	s/:0xcr[0-7],/:0x/
+	s/:cr[0-7],/:/
 	p
 }' | {
 
 all_good=true
 while IFS=: read -r from branch to sym; do
+	case "$to" in
+	c*)	to="0x$to"
+		;;
+	.+*)
+		to=${to#.+}
+		if [ "$branch" = 'b' ]; then
+			if (( to >= 0x2000000 )); then
+				to=$(( to - 0x4000000 ))
+			fi
+		elif (( to >= 0x8000 )); then
+			to=$(( to - 0x10000 ))
+		fi
+		printf -v to '0x%x' $(( "0x$from" + to ))
+		;;
+	*)	printf 'Unkown branch format\n'
+		;;
+	esac
+	if [ "$to" = "$sim" ]; then
+		continue
+	fi
 	if (( to > end_intr )); then
 		if $all_good; then
 			printf '%s\n' 'WARNING: Unrelocated relative branches'
-- 
2.28.0



More information about the Linuxppc-dev mailing list