Merge tag 'objtool-core-2024-07-16' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip

Pull objtool updates from Ingo Molnar:

 - Fix bug that caused objtool to confuse certain memory ops added by
   KASAN instrumentation as stack accesses

 - Various faddr2line optimizations

 - Improve error messages

* tag 'objtool-core-2024-07-16' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  objtool/x86: objtool can confuse memory and stack access
  objtool: Use "action" in error message to be consistent with help
  scripts/faddr2line: Check only two symbols when calculating symbol size
  scripts/faddr2line: Remove call to addr2line from find_dir_prefix()
  scripts/faddr2line: Invoke addr2line as a single long-running process
  scripts/faddr2line: Pass --addresses argument to addr2line
  scripts/faddr2line: Check vmlinux only once
  scripts/faddr2line: Combine three readelf calls into one
  scripts/faddr2line: Reduce number of readelf calls to three

scripts/faddr2line

@@ -85,15 +85,17 @@ command -v ${ADDR2LINE} >/dev/null 2>&1 || die "${ADDR2LINE} isn't installed"
 # init/main.c!  This only works for vmlinux.  Otherwise it falls back to
 # printing the absolute path.
 find_dir_prefix() {
-	local objfile=$1
-
-	local start_kernel_addr=$(${READELF} --symbols --wide $objfile | sed 's/\[.*\]//' |
+	local start_kernel_addr=$(echo "${ELF_SYMS}" | sed 's/\[.*\]//' |
 		${AWK} '$8 == "start_kernel" {printf "0x%s", $2}')
 	[[ -z $start_kernel_addr ]] && return
 
-	local file_line=$(${ADDR2LINE} -e $objfile $start_kernel_addr)
-	[[ -z $file_line ]] && return
+	run_addr2line ${start_kernel_addr} ""
+	[[ -z $ADDR2LINE_OUT ]] && return
 
+	local file_line=${ADDR2LINE_OUT#* at }
+	if [[ -z $file_line ]] || [[ $file_line = $ADDR2LINE_OUT ]]; then
+		return
+	fi
 	local prefix=${file_line%init/main.c:*}
 	if [[ -z $prefix ]] || [[ $prefix = $file_line ]]; then
 		return
@@ -103,6 +105,71 @@ find_dir_prefix() {
 	return 0
 }
 
+run_readelf() {
+	local objfile=$1
+	local out=$(${READELF} --file-header --section-headers --symbols --wide $objfile)
+
+	# This assumes that readelf first prints the file header, then the section headers, then the symbols.
+	# Note: It seems that GNU readelf does not prefix section headers with the "There are X section headers"
+	# line when multiple options are given, so let's also match with the "Section Headers:" line.
+	ELF_FILEHEADER=$(echo "${out}" | sed -n '/There are [0-9]* section headers, starting at offset\|Section Headers:/q;p')
+	ELF_SECHEADERS=$(echo "${out}" | sed -n '/There are [0-9]* section headers, starting at offset\|Section Headers:/,$p' | sed -n '/Symbol table .* contains [0-9]* entries:/q;p')
+	ELF_SYMS=$(echo "${out}" | sed -n '/Symbol table .* contains [0-9]* entries:/,$p')
+}
+
+check_vmlinux() {
+	# vmlinux uses absolute addresses in the section table rather than
+	# section offsets.
+	IS_VMLINUX=0
+	local file_type=$(echo "${ELF_FILEHEADER}" |
+		${AWK} '$1 == "Type:" { print $2; exit }')
+	if [[ $file_type = "EXEC" ]] || [[ $file_type == "DYN" ]]; then
+		IS_VMLINUX=1
+	fi
+}
+
+init_addr2line() {
+	local objfile=$1
+
+	check_vmlinux
+
+	ADDR2LINE_ARGS="--functions --pretty-print --inlines --addresses --exe=$objfile"
+	if [[ $IS_VMLINUX = 1 ]]; then
+		# If the executable file is vmlinux, we don't pass section names to
+		# addr2line, so we can launch it now as a single long-running process.
+		coproc ADDR2LINE_PROC (${ADDR2LINE} ${ADDR2LINE_ARGS})
+	fi
+}
+
+run_addr2line() {
+	local addr=$1
+	local sec_name=$2
+
+	if [[ $IS_VMLINUX = 1 ]]; then
+		# We send to the addr2line process: (1) the address, then (2) a sentinel
+		# value, i.e., something that can't be interpreted as a valid address
+		# (i.e., ","). This causes addr2line to write out: (1) the answer for
+		# our address, then (2) either "?? ??:0" or "0x0...0: ..." (if
+		# using binutils' addr2line), or "," (if using LLVM's addr2line).
+		echo ${addr} >& "${ADDR2LINE_PROC[1]}"
+		echo "," >& "${ADDR2LINE_PROC[1]}"
+		local first_line
+		read -r first_line <& "${ADDR2LINE_PROC[0]}"
+		ADDR2LINE_OUT=$(echo "${first_line}" | sed 's/^0x[0-9a-fA-F]*: //')
+		while read -r line <& "${ADDR2LINE_PROC[0]}"; do
+			if [[ "$line" == "?? ??:0" ]] || [[ "$line" == "," ]] || [[ $(echo "$line" | ${GREP} "^0x00*: ") ]]; then
+				break
+			fi
+			ADDR2LINE_OUT+=$'\n'$(echo "$line" | sed 's/^0x[0-9a-fA-F]*: //')
+		done
+	else
+		# Run addr2line as a single invocation.
+		local sec_arg
+		[[ -z $sec_name ]] && sec_arg="" || sec_arg="--section=${sec_name}"
+		ADDR2LINE_OUT=$(${ADDR2LINE} ${ADDR2LINE_ARGS} ${sec_arg} ${addr} | sed 's/^0x[0-9a-fA-F]*: //')
+	fi
+}
+
 __faddr2line() {
 	local objfile=$1
 	local func_addr=$2
@@ -113,8 +180,6 @@ __faddr2line() {
 	local func_offset=${func_addr#*+}
 	func_offset=${func_offset%/*}
 	local user_size=
-	local file_type
-	local is_vmlinux=0
 	[[ $func_addr =~ "/" ]] && user_size=${func_addr#*/}
 
 	if [[ -z $sym_name ]] || [[ -z $func_offset ]] || [[ $sym_name = $func_addr ]]; then
@@ -123,14 +188,6 @@ __faddr2line() {
 		return
 	fi
 
-	# vmlinux uses absolute addresses in the section table rather than
-	# section offsets.
-	local file_type=$(${READELF} --file-header $objfile |
-		${AWK} '$1 == "Type:" { print $2; exit }')
-	if [[ $file_type = "EXEC" ]] || [[ $file_type == "DYN" ]]; then
-		is_vmlinux=1
-	fi
-
 	# Go through each of the object's symbols which match the func name.
 	# In rare cases there might be duplicates, in which case we print all
 	# matches.
@@ -143,8 +200,7 @@ __faddr2line() {
 		local sec_name
 
 		# Get the section size:
-		sec_size=$(${READELF} --section-headers --wide $objfile |
-			sed 's/\[ /\[/' |
+		sec_size=$(echo "${ELF_SECHEADERS}" | sed 's/\[ /\[/' |
 			${AWK} -v sec=$sym_sec '$1 == "[" sec "]" { print "0x" $6; exit }')
 
 		if [[ -z $sec_size ]]; then
@@ -154,8 +210,7 @@ __faddr2line() {
 		fi
 
 		# Get the section name:
-		sec_name=$(${READELF} --section-headers --wide $objfile |
-			sed 's/\[ /\[/' |
+		sec_name=$(echo "${ELF_SECHEADERS}" | sed 's/\[ /\[/' |
 			${AWK} -v sec=$sym_sec '$1 == "[" sec "]" { print $2; exit }')
 
 		if [[ -z $sec_name ]]; then
@@ -197,7 +252,7 @@ __faddr2line() {
 				found=2
 				break
 			fi
-		done < <(${READELF} --symbols --wide $objfile | sed 's/\[.*\]//' | ${AWK} -v sec=$sym_sec '$7 == sec' | sort --key=2)
+		done < <(echo "${ELF_SYMS}" | sed 's/\[.*\]//' | ${AWK} -v sec=$sym_sec '$7 == sec' | sort --key=2 | ${GREP} -A1 --no-group-separator " ${sym_name}$")
 
 		if [[ $found = 0 ]]; then
 			warn "can't find symbol: sym_name: $sym_name sym_sec: $sym_sec sym_addr: $sym_addr sym_elf_size: $sym_elf_size"
@@ -249,9 +304,8 @@ __faddr2line() {
 
 		# Pass section address to addr2line and strip absolute paths
 		# from the output:
-		local args="--functions --pretty-print --inlines --exe=$objfile"
-		[[ $is_vmlinux = 0 ]] && args="$args --section=$sec_name"
-		local output=$(${ADDR2LINE} $args $addr | sed "s; $dir_prefix\(\./\)*; ;")
+		run_addr2line $addr $sec_name
+		local output=$(echo "${ADDR2LINE_OUT}" | sed "s; $dir_prefix\(\./\)*; ;")
 		[[ -z $output ]] && continue
 
 		# Default output (non --list):
@@ -278,7 +332,7 @@ __faddr2line() {
 
 		DONE=1
 
-	done < <(${READELF} --symbols --wide $objfile | sed 's/\[.*\]//' | ${AWK} -v fn=$sym_name '$8 == fn')
+	done < <(echo "${ELF_SYMS}" | sed 's/\[.*\]//' | ${AWK} -v fn=$sym_name '$8 == fn')
 }
 
 [[ $# -lt 2 ]] && usage
@@ -291,10 +345,14 @@ LIST=0
 [[ ! -f $objfile ]] && die "can't find objfile $objfile"
 shift
 
-${READELF} --section-headers --wide $objfile | ${GREP} -q '\.debug_info' || die "CONFIG_DEBUG_INFO not enabled"
+run_readelf $objfile
+
+echo "${ELF_SECHEADERS}" | ${GREP} -q '\.debug_info' || die "CONFIG_DEBUG_INFO not enabled"
+
+init_addr2line $objfile
 
 DIR_PREFIX=supercalifragilisticexpialidocious
-find_dir_prefix $objfile
+find_dir_prefix
 
 FIRST=1
 while [[ $# -gt 0 ]]; do

tools/objtool/arch/x86/decode.c

@@ -125,8 +125,14 @@ bool arch_pc_relative_reloc(struct reloc *reloc)
 #define is_RIP()   ((modrm_rm & 7) == CFI_BP && modrm_mod == 0)
 #define have_SIB() ((modrm_rm & 7) == CFI_SP && mod_is_mem())
 
+/*
+ * Check the ModRM register. If there is a SIB byte then check with
+ * the SIB base register. But if the SIB base is 5 (i.e. CFI_BP) and
+ * ModRM mod is 0 then there is no base register.
+ */
 #define rm_is(reg) (have_SIB() ? \
-		    sib_base == (reg) && sib_index == CFI_SP : \
+		    sib_base == (reg) && sib_index == CFI_SP && \
+		    (sib_base != CFI_BP || modrm_mod != 0) :	\
 		    modrm_rm == (reg))
 
 #define rm_is_mem(reg)	(mod_is_mem() && !is_RIP() && rm_is(reg))

tools/objtool/builtin-check.c

@@ -144,7 +144,7 @@ static bool opts_valid(void)
 	    opts.static_call		||
 	    opts.uaccess) {
 		if (opts.dump_orc) {
-			ERROR("--dump can't be combined with other options");
+			ERROR("--dump can't be combined with other actions");
 			return false;
 		}
 
@@ -159,7 +159,7 @@ static bool opts_valid(void)
 	if (opts.dump_orc)
 		return true;
 
-	ERROR("At least one command required");
+	ERROR("At least one action required");
 	return false;
 }