[UpdateTestChecks] Handle prefix reuse for appended check lines

[jdoerfert]

Summary

When we appended check lines at the end we could not share prefixes
before. This patch should make it possible and allow us to reduce
some check line counts (especially for Clang/OpenMP tests).

See also: https://reviews.llvm.org/D128686

[mtrofin]

nit: in the commit message, there's a spurious 'the' before 'some check counts'

[jhuber6]

LGTM, others may want to comment.

llvm/test/tools/UpdateTestChecks/update_test_checks/Inputs/generated_funcs_prefix_reuse.ll

+; RUN: opt -passes=hotcoldsplit -hotcoldsplit-threshold=0 -S < %s | FileCheck %s --check-prefix=REUSE
+; RUN: opt -passes=hotcoldsplit -hotcoldsplit-threshold=0 -S < %s | FileCheck %s --check-prefix=REUSE
+target datalayout = "e-m:o-i64:64-f80:128-n8:16:32:64-S128"
+target triple = "x86_64-apple-macosx10.14.0"
+
+@G = external dso_local global i32, align 4
+
+define void @foo(i32) {
+  %2 = icmp eq i32 %0, 0
+  tail call void @_Z10sideeffectv()
+  br i1 %2, label %sink, label %exit
+
+sink:
+  tail call void @_Z10sideeffectv()
+  call void @llvm.trap()
+  unreachable
+
+exit:
+  ret void
+}
+
+define void @bar(i32) {
+  %2 = icmp eq i32 %0, 0
+  tail call void @_Z10sideeffectv()
+  br i1 %2, label %sink, label %exit
+
+sink:
+  tail call void @_Z10sideeffectv()
+  call void @llvm.trap()
+  unreachable
+
+exit:
+  ret void
+}
+
+declare void @llvm.trap() noreturn cold
+declare void @_Z10sideeffectv()

llvm/test/tools/UpdateTestChecks/update_test_checks/Inputs/generated_funcs_prefix_reuse.ll.generated.expected

+; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --include-generated-funcs
+; RUN: opt -passes=hotcoldsplit -hotcoldsplit-threshold=0 -S < %s | FileCheck %s --check-prefix=REUSE
+; RUN: opt -passes=hotcoldsplit -hotcoldsplit-threshold=0 -S < %s | FileCheck %s --check-prefix=REUSE
+target datalayout = "e-m:o-i64:64-f80:128-n8:16:32:64-S128"
+target triple = "x86_64-apple-macosx10.14.0"
+
+@G = external dso_local global i32, align 4
+
+define void @foo(i32) {
+  %2 = icmp eq i32 %0, 0
+  tail call void @_Z10sideeffectv()
+  br i1 %2, label %sink, label %exit
+
+sink:
+  tail call void @_Z10sideeffectv()
+  call void @llvm.trap()
+  unreachable
+
+exit:
+  ret void
+}
+
+define void @bar(i32) {
+  %2 = icmp eq i32 %0, 0
+  tail call void @_Z10sideeffectv()
+  br i1 %2, label %sink, label %exit
+
+sink:
+  tail call void @_Z10sideeffectv()
+  call void @llvm.trap()
+  unreachable
+
+exit:
+  ret void
+}
+
+declare void @llvm.trap() noreturn cold
+declare void @_Z10sideeffectv()
+; REUSE-LABEL: @foo(
+; REUSE-NEXT:    [[TMP2:%.*]] = icmp eq i32 [[TMP0:%.*]], 0
+; REUSE-NEXT:    tail call void @_Z10sideeffectv()
+; REUSE-NEXT:    br i1 [[TMP2]], label [[CODEREPL:%.*]], label [[EXIT:%.*]]
+; REUSE:       codeRepl:
+; REUSE-NEXT:    call void @foo.cold.1() #[[ATTR2:[0-9]+]]
+; REUSE-NEXT:    ret void
+; REUSE:       exit:
+; REUSE-NEXT:    ret void
+;
+;
+; REUSE-LABEL: @bar(
+; REUSE-NEXT:    [[TMP2:%.*]] = icmp eq i32 [[TMP0:%.*]], 0
+; REUSE-NEXT:    tail call void @_Z10sideeffectv()
+; REUSE-NEXT:    br i1 [[TMP2]], label [[CODEREPL:%.*]], label [[EXIT:%.*]]
+; REUSE:       codeRepl:
+; REUSE-NEXT:    call void @bar.cold.1() #[[ATTR2]]
+; REUSE-NEXT:    ret void
+; REUSE:       exit:
+; REUSE-NEXT:    ret void
+;
+;
+; REUSE-LABEL: @foo.cold.1(
+; REUSE-NEXT:  newFuncRoot:
+; REUSE-NEXT:    br label [[SINK:%.*]]
+; REUSE:       sink:
+; REUSE-NEXT:    tail call void @_Z10sideeffectv()
+; REUSE-NEXT:    call void @llvm.trap()
+; REUSE-NEXT:    unreachable
+;
+;
+; REUSE-LABEL: @bar.cold.1(
+; REUSE-NEXT:  newFuncRoot:
+; REUSE-NEXT:    br label [[SINK:%.*]]
+; REUSE:       sink:
+; REUSE-NEXT:    tail call void @_Z10sideeffectv()
+; REUSE-NEXT:    call void @llvm.trap()
+; REUSE-NEXT:    unreachable
+;

llvm/test/tools/UpdateTestChecks/update_test_checks/Inputs/generated_funcs_prefix_reuse.ll.generated.globals.expected

+; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --check-globals --include-generated-funcs
+; RUN: opt -passes=hotcoldsplit -hotcoldsplit-threshold=0 -S < %s | FileCheck %s --check-prefix=REUSE
+; RUN: opt -passes=hotcoldsplit -hotcoldsplit-threshold=0 -S < %s | FileCheck %s --check-prefix=REUSE
+target datalayout = "e-m:o-i64:64-f80:128-n8:16:32:64-S128"
+target triple = "x86_64-apple-macosx10.14.0"
+
+@G = external dso_local global i32, align 4
+
+define void @foo(i32) {
+  %2 = icmp eq i32 %0, 0
+  tail call void @_Z10sideeffectv()
+  br i1 %2, label %sink, label %exit
+
+sink:
+  tail call void @_Z10sideeffectv()
+  call void @llvm.trap()
+  unreachable
+
+exit:
+  ret void
+}
+
+define void @bar(i32) {
+  %2 = icmp eq i32 %0, 0
+  tail call void @_Z10sideeffectv()
+  br i1 %2, label %sink, label %exit
+
+sink:
+  tail call void @_Z10sideeffectv()
+  call void @llvm.trap()
+  unreachable
+
+exit:
+  ret void
+}
+
+declare void @llvm.trap() noreturn cold
+declare void @_Z10sideeffectv()
+;.
+; REUSE: @[[G:[a-zA-Z0-9_$"\\.-]+]] = external dso_local global i32, align 4
+;.
+; REUSE-LABEL: @foo(
+; REUSE-NEXT:    [[TMP2:%.*]] = icmp eq i32 [[TMP0:%.*]], 0
+; REUSE-NEXT:    tail call void @_Z10sideeffectv()
+; REUSE-NEXT:    br i1 [[TMP2]], label [[CODEREPL:%.*]], label [[EXIT:%.*]]
+; REUSE:       codeRepl:
+; REUSE-NEXT:    call void @foo.cold.1() #[[ATTR2:[0-9]+]]
+; REUSE-NEXT:    ret void
+; REUSE:       exit:
+; REUSE-NEXT:    ret void
+;
+;
+; REUSE-LABEL: @bar(
+; REUSE-NEXT:    [[TMP2:%.*]] = icmp eq i32 [[TMP0:%.*]], 0
+; REUSE-NEXT:    tail call void @_Z10sideeffectv()
+; REUSE-NEXT:    br i1 [[TMP2]], label [[CODEREPL:%.*]], label [[EXIT:%.*]]
+; REUSE:       codeRepl:
+; REUSE-NEXT:    call void @bar.cold.1() #[[ATTR2]]
+; REUSE-NEXT:    ret void
+; REUSE:       exit:
+; REUSE-NEXT:    ret void
+;
+;
+; REUSE-LABEL: @foo.cold.1(
+; REUSE-NEXT:  newFuncRoot:
+; REUSE-NEXT:    br label [[SINK:%.*]]
+; REUSE:       sink:
+; REUSE-NEXT:    tail call void @_Z10sideeffectv()
+; REUSE-NEXT:    call void @llvm.trap()
+; REUSE-NEXT:    unreachable
+;
+;
+; REUSE-LABEL: @bar.cold.1(
+; REUSE-NEXT:  newFuncRoot:
+; REUSE-NEXT:    br label [[SINK:%.*]]
+; REUSE:       sink:
+; REUSE-NEXT:    tail call void @_Z10sideeffectv()
+; REUSE-NEXT:    call void @llvm.trap()
+; REUSE-NEXT:    unreachable
+;
+;.
+; REUSE: attributes #[[ATTR0:[0-9]+]] = { cold noreturn nounwind }
+; REUSE: attributes #[[ATTR1:[0-9]+]] = { cold minsize noreturn }
+; REUSE: attributes #[[ATTR2]] = { noinline }
+;.

llvm/test/tools/UpdateTestChecks/update_test_checks/Inputs/generated_funcs_prefix_reuse.ll.nogenerated.expected

+; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
+; RUN: opt -passes=hotcoldsplit -hotcoldsplit-threshold=0 -S < %s | FileCheck %s --check-prefix=REUSE
+; RUN: opt -passes=hotcoldsplit -hotcoldsplit-threshold=0 -S < %s | FileCheck %s --check-prefix=REUSE
+target datalayout = "e-m:o-i64:64-f80:128-n8:16:32:64-S128"
+target triple = "x86_64-apple-macosx10.14.0"
+
+@G = external dso_local global i32, align 4
+
+define void @foo(i32) {
+; REUSE-LABEL: @foo(
+; REUSE-NEXT:    [[TMP2:%.*]] = icmp eq i32 [[TMP0:%.*]], 0
+; REUSE-NEXT:    tail call void @_Z10sideeffectv()
+; REUSE-NEXT:    br i1 [[TMP2]], label [[CODEREPL:%.*]], label [[EXIT:%.*]]
+; REUSE:       codeRepl:
+; REUSE-NEXT:    call void @foo.cold.1() #[[ATTR2:[0-9]+]]
+; REUSE-NEXT:    ret void
+; REUSE:       exit:
+; REUSE-NEXT:    ret void
+;
+  %2 = icmp eq i32 %0, 0
+  tail call void @_Z10sideeffectv()
+  br i1 %2, label %sink, label %exit
+
+sink:
+  tail call void @_Z10sideeffectv()
+  call void @llvm.trap()
+  unreachable
+
+exit:
+  ret void
+}
+
+define void @bar(i32) {
+; REUSE-LABEL: @bar(
+; REUSE-NEXT:    [[TMP2:%.*]] = icmp eq i32 [[TMP0:%.*]], 0
+; REUSE-NEXT:    tail call void @_Z10sideeffectv()
+; REUSE-NEXT:    br i1 [[TMP2]], label [[CODEREPL:%.*]], label [[EXIT:%.*]]
+; REUSE:       codeRepl:
+; REUSE-NEXT:    call void @bar.cold.1() #[[ATTR2]]
+; REUSE-NEXT:    ret void
+; REUSE:       exit:
+; REUSE-NEXT:    ret void
+;
+  %2 = icmp eq i32 %0, 0
+  tail call void @_Z10sideeffectv()
+  br i1 %2, label %sink, label %exit
+
+sink:
+  tail call void @_Z10sideeffectv()
+  call void @llvm.trap()
+  unreachable
+
+exit:
+  ret void
+}
+
+declare void @llvm.trap() noreturn cold
+declare void @_Z10sideeffectv()

llvm/test/tools/UpdateTestChecks/update_test_checks/Inputs/generated_funcs_prefix_reuse.ll.nogenerated.globals.expected

+; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --check-globals
+; RUN: opt -passes=hotcoldsplit -hotcoldsplit-threshold=0 -S < %s | FileCheck %s --check-prefix=REUSE
+; RUN: opt -passes=hotcoldsplit -hotcoldsplit-threshold=0 -S < %s | FileCheck %s --check-prefix=REUSE
+target datalayout = "e-m:o-i64:64-f80:128-n8:16:32:64-S128"
+target triple = "x86_64-apple-macosx10.14.0"
+
+@G = external dso_local global i32, align 4
+
+;.
+; REUSE: @[[G:[a-zA-Z0-9_$"\\.-]+]] = external dso_local global i32, align 4
+;.
+define void @foo(i32) {
+; REUSE-LABEL: @foo(
+; REUSE-NEXT:    [[TMP2:%.*]] = icmp eq i32 [[TMP0:%.*]], 0
+; REUSE-NEXT:    tail call void @_Z10sideeffectv()
+; REUSE-NEXT:    br i1 [[TMP2]], label [[CODEREPL:%.*]], label [[EXIT:%.*]]
+; REUSE:       codeRepl:
+; REUSE-NEXT:    call void @foo.cold.1() #[[ATTR2:[0-9]+]]
+; REUSE-NEXT:    ret void
+; REUSE:       exit:
+; REUSE-NEXT:    ret void
+;
+  %2 = icmp eq i32 %0, 0
+  tail call void @_Z10sideeffectv()
+  br i1 %2, label %sink, label %exit
+
+sink:
+  tail call void @_Z10sideeffectv()
+  call void @llvm.trap()
+  unreachable
+
+exit:
+  ret void
+}
+
+define void @bar(i32) {
+; REUSE-LABEL: @bar(
+; REUSE-NEXT:    [[TMP2:%.*]] = icmp eq i32 [[TMP0:%.*]], 0
+; REUSE-NEXT:    tail call void @_Z10sideeffectv()
+; REUSE-NEXT:    br i1 [[TMP2]], label [[CODEREPL:%.*]], label [[EXIT:%.*]]
+; REUSE:       codeRepl:
+; REUSE-NEXT:    call void @bar.cold.1() #[[ATTR2]]
+; REUSE-NEXT:    ret void
+; REUSE:       exit:
+; REUSE-NEXT:    ret void
+;
+  %2 = icmp eq i32 %0, 0
+  tail call void @_Z10sideeffectv()
+  br i1 %2, label %sink, label %exit
+
+sink:
+  tail call void @_Z10sideeffectv()
+  call void @llvm.trap()
+  unreachable
+
+exit:
+  ret void
+}
+
+declare void @llvm.trap() noreturn cold
+declare void @_Z10sideeffectv()
+;.
+; REUSE: attributes #[[ATTR0:[0-9]+]] = { cold noreturn nounwind }
+; REUSE: attributes #[[ATTR1:[0-9]+]] = { cold minsize noreturn }
+; REUSE: attributes #[[ATTR2]] = { noinline }
+;.

llvm/test/tools/UpdateTestChecks/update_test_checks/generated_funcs_prefix_reuse.test

+## Check that generated functions are included.
+# RUN: cp -f %S/Inputs/generated_funcs_prefix_reuse.ll %t.ll && %update_test_checks --include-generated-funcs %t.ll
+# RUN: diff -u %t.ll %S/Inputs/generated_funcs_prefix_reuse.ll.generated.expected
+
+## Check that running the script again does not change the result:
+# RUN: %update_test_checks --include-generated-funcs %t.ll
+# RUN: diff -u %t.ll %S/Inputs/generated_funcs_prefix_reuse.ll.generated.expected
+## Also try the --check-globals flag
+# RUN: %update_test_checks %t.ll --check-globals
+# RUN: diff -u %t.ll %S/Inputs/generated_funcs_prefix_reuse.ll.generated.globals.expected
+
+## Check that generated functions are not included.
+# RUN: cp -f %S/Inputs/generated_funcs_prefix_reuse.ll %t.ll && %update_test_checks %t.ll
+# RUN: diff -u %t.ll %S/Inputs/generated_funcs_prefix_reuse.ll.nogenerated.expected
+
+## Check that running the script again does not change the result:
+# RUN: %update_test_checks %t.ll
+# RUN: diff -u %t.ll %S/Inputs/generated_funcs_prefix_reuse.ll.nogenerated.expected
+## Also try the --check-globals flag
+# RUN: %update_test_checks %t.ll --check-globals
+# RUN: diff -u %t.ll %S/Inputs/generated_funcs_prefix_reuse.ll.nogenerated.globals.expected

llvm/utils/UpdateTestChecks/common.py

if line.lstrip().startswith(comment_string):
       m = CHECK_RE.match(line)
       if m and m.group(1) in prefix_set:
         continue
     output_lines.append(line.rstrip('\n'))
 
 def add_checks_at_end(output_lines, prefix_list, func_order,
                       comment_string, check_generator):
   added = set()
-  generated_prefixes = []
+  generated_prefixes = set()
   for prefix in prefix_list:
     prefixes = prefix[0]
     tool_args = prefix[1]
     for prefix in prefixes:
       for func in func_order[prefix]:
+        # The func order can contain the same functions multiple times.
+        # If we see one again we are done.
+        if (func, prefix) in added:
+            continue
         if added:
           output_lines.append(comment_string)
-        added.add(func)
 
         # The add_*_checks routines expect a run list whose items are
         # tuples that have a list of prefixes as their first element and
         # tool command args string as their second element.  They output
         # checks for each prefix in the list of prefixes.  By doing so, it
         # implicitly assumes that for each function every run line will
         # generate something for that function.  That is not the case for
         # generated functions as some run lines might not generate them
         # (e.g. -fopenmp vs. no -fopenmp).
         #
         # Therefore, pass just the prefix we're interested in.  This has
         # the effect of generating all of the checks for functions of a
         # single prefix before moving on to the next prefix.  So checks
         # are ordered by prefix instead of by function as in "normal"
         # mode.
-        generated_prefixes.extend(check_generator(output_lines,
-                        [([prefix], tool_args)],
-                        func))
+        for generated_prefix in check_generator(output_lines,
+                        [([prefix], tool_args)], func):
+            added.add((func, generated_prefix))
+            generated_prefixes.add(generated_prefix)
   return generated_prefixes