2 files changed, 32 insertions, 3 deletions

diff --git a/gcc/testsuite/lib/scanasm.exp b/gcc/testsuite/lib/scanasm.exp
index 4b018ab..fb53544 100644
--- a/gcc/testsuite/lib/scanasm.exp
+++ b/gcc/testsuite/lib/scanasm.exp
@@ -24,19 +24,33 @@ proc make_pattern_printable { pattern } {
     return [string map {\t \\t \n \\n \r \\r \\ \\\\} $pattern]
 }
 
+# Append to ARGS to make it suitable for use by dg-scan to indicate
+# that encoding ENC should be used when reading from the file.
+
+proc append_encoding_arg { args enc } {
+    if { [llength $args] < 2 } {
+	# Add target selector.
+	lappend args { target "*-*-*" }
+    }
+    # Add encoding ENC.
+    lappend args $enc
+    return $args
+}
+
 # Scan the OUTPUT_FILE for a pattern.  If it is present and POSITIVE
 # is non-zero, or it is not present and POSITIVE is zero, the test
 # passes.  The ORIG_ARGS is the list of arguments provided by dg-final
 # to scan-assembler.  The first element in ORIG_ARGS is the regular
 # expression to look for in the file.  The second element, if present,
-# is a DejaGNU target selector.
+# is a DejaGNU target selector.  The third element, if present, is the
+# encoding to use when reading from the file.
 
 proc dg-scan { name positive testcase output_file orig_args } {
     if { [llength $orig_args] < 1 } {
 	error "$name: too few arguments"
         return
     }
-    if { [llength $orig_args] > 2 } {
+    if { [llength $orig_args] > 3 } {
 	error "$name: too many arguments"
 	return
     }
@@ -59,6 +73,10 @@ proc dg-scan { name positive testcase output_file orig_args } {
 	return
     }
     set fd [open $output_file r]
+    if { [llength $orig_args] >= 3 } {
+	set file_encoding [lindex $orig_args 2]
+	fconfigure $fd -encoding $file_encoding
+    }
     set text [read $fd]
     close $fd
 
diff --git a/gcc/testsuite/lib/scansarif.exp b/gcc/testsuite/lib/scansarif.exp
index d06390b..05524aa 100644
--- a/gcc/testsuite/lib/scansarif.exp
+++ b/gcc/testsuite/lib/scansarif.exp
@@ -17,7 +17,10 @@
 # Various utilities for scanning SARIF output, used by gcc-dg.exp and
 # g++-dg.exp.
 #
-# This is largely borrowed from scanasm.exp.
+# This is largely borrowed from scanasm.exp, but tweaked to force Tcl
+# to treat the file as UTF-8: section 3.1 of SARIF 2.1.0
+# ("File Format" > "General") specifies: "A SARIF log file SHALL be
+# encoded in UTF-8 [RFC3629])".
 
 # Look for a pattern in the .sarif file produced by the compiler.  See
 # dg-scan for details.
@@ -27,6 +30,10 @@ proc scan-sarif-file { args } {
     # The name might include a list of options; extract the file name.
     set filename [lindex $testcase 0]
     set output_file "[file tail $filename].sarif"
+
+    # Treat the file as UTF-8 encoded when reading it.
+    set args [append_encoding_arg $args "utf-8"]
+
     dg-scan "scan-sarif-file" 1 $testcase $output_file $args
 }
 
@@ -38,5 +45,9 @@ proc scan-sarif-file-not { args } {
     # The name might include a list of options; extract the file name.
     set filename [lindex $testcase 0]
     set output_file "[file tail $filename].sarif"
+
+    # Treat the file as UTF-8 encoded when reading it.
+    set args [append_encoding_arg $args "utf-8"]
+
     dg-scan "scan-sarif-file-not" 0 $testcase $output_file $args
 }