[llvm] [ThinLTO] Allow importing based on a workload definition (PR #74545)

[Teresa Johnson via llvm-commits]

================
@@ -1,67 +1,153 @@
+; Set up
+; RUN: rm -rf %t
+; RUN: mkdir -p %t
+; RUN: split-file %s %t
+;
+;--- m1.ll
+target datalayout = "e-m:e-p270:32:32-p271:32:32-p272:64:64-i64:64-f80:128-n8:16:32:64-S128"
+target triple = "x86_64-pc-linux-gnu"
+
+declare void @m1_variant()
+declare void @m2_f1_alias()
+
+define dso_local void @m1_f1() {
+  call void @m1_f2()
+  call void @noninterposable_f()
+  ret void
+}
+
+define internal void @m1_f2() {
+  call void @interposable_f()
+  ret void
+}
+
+define external void @interposable_f() {
+  call void @m1_variant()
+  ret void
+}
+
+define linkonce_odr void @noninterposable_f() {
+  call void @m1_variant()
+  ret void
+}
+;--- m2.ll
+target datalayout = "e-m:e-p270:32:32-p271:32:32-p272:64:64-i64:64-f80:128-n8:16:32:64-S128"
+target triple = "x86_64-pc-linux-gnu"
+
+declare void @m2_variant()
+
+define dso_local void @m2_f1() {
+  call void @interposable_f()
+  call void @noninterposable_f()
+  ret void
+}
+
+ at m2_f1_alias = alias void (...), ptr @m2_f1
+
+define weak void @interposable_f() {
+  call void @m2_variant() 
+  ret void
+}
+
+define linkonce_odr void @noninterposable_f() {
+  call void @m2_variant()
+  ret void
+}
+;--- m3.ll
+target datalayout = "e-m:e-p270:32:32-p271:32:32-p272:64:64-i64:64-f80:128-n8:16:32:64-S128"
+target triple = "x86_64-pc-linux-gnu"
+
+declare void @m1_f1()
+
+define dso_local void @m3_f1() {
+  call void @m1_f1()
+  ret void
+}
+;
+; RUN: opt -module-summary %t/m1.ll -o %t/m1.bc
+; RUN: opt -module-summary %t/m2.ll -o %t/m2.bc
+; RUN: opt -module-summary %t/m3.ll -o %t/m3.bc
+; RUN: rm -rf %t_baseline
+; RUN: rm -rf %t_exp
 ; RUN: mkdir -p %t_baseline
-; RUN: opt -module-summary %S/Inputs/workload1.ll -o %t_baseline/workload1.bc
-; RUN: opt -module-summary %S/Inputs/workload2.ll -o %t_baseline/workload2.bc
-; RUN: opt -module-summary %S/Inputs/workload3.ll -o %t_baseline/workload3.bc
+; RUN: mkdir -p %t_exp
 ;
-; Normal run. The first module shouldn't get m2_f1
+; Normal run. m1 shouldn't get m2_f1 because it's not referenced from there.
 ;
-; RUN: llvm-lto2 run %t_baseline/workload1.bc %t_baseline/workload2.bc %t_baseline/workload3.bc \
+; RUN: llvm-lto2 run %t/m1.bc %t/m2.bc %t/m3.bc \
 ; RUN:  -o %t_baseline/result.o -save-temps \
-; RUN:  -r %t_baseline/workload1.bc,m1_f1,plx \
-; RUN:  -r %t_baseline/workload1.bc,interposable_f \
-; RUN:  -r %t_baseline/workload1.bc,noninterposable_f \
-; RUN:  -r %t_baseline/workload1.bc,m1_variant \
-; RUN:  -r %t_baseline/workload2.bc,m2_f1,plx \
-; RUN:  -r %t_baseline/workload2.bc,m2_f1_alias \
-; RUN:  -r %t_baseline/workload2.bc,interposable_f,p \
-; RUN:  -r %t_baseline/workload2.bc,noninterposable_f,p \
-; RUN:  -r %t_baseline/workload2.bc,m2_variant \
-; RUN:  -r %t_baseline/workload3.bc,m1_f1 \
-; RUN:  -r %t_baseline/workload3.bc,m3_f1,plx
+; RUN:  -r %t/m1.bc,m1_f1,plx \
+; RUN:  -r %t/m1.bc,interposable_f,p \
+; RUN:  -r %t/m1.bc,noninterposable_f \
+; RUN:  -r %t/m1.bc,m1_variant \
+; RUN:  -r %t/m1.bc,m2_f1_alias \
+; RUN:  -r %t/m2.bc,m2_f1,plx \
+; RUN:  -r %t/m2.bc,m2_f1_alias,plx \
+; RUN:  -r %t/m2.bc,interposable_f \
+; RUN:  -r %t/m2.bc,noninterposable_f,p \
+; RUN:  -r %t/m2.bc,m2_variant \
+; RUN:  -r %t/m3.bc,m1_f1 \
+; RUN:  -r %t/m3.bc,m3_f1,plx
 ; RUN: llvm-dis %t_baseline/result.o.1.3.import.bc -o - | FileCheck %s --check-prefix=NOPROF
 ;
-; NOPROF-NOT: m2_f1
+; NOPROF-NOT: m2_f1()
 ;
 ; The run with workload definitions - same other options.
 ;
-; RUN: mkdir -p %t
-; RUN: opt -module-summary %S/Inputs/workload1.ll -o %t/workload1.bc
-; RUN: opt -module-summary %S/Inputs/workload2.ll -o %t/workload2.bc
-; RUN: opt -module-summary %S/Inputs/workload3.ll -o %t/workload3.bc
-; RUN: echo '{"m1_f1":["m2_f1", "m2_f1_alias", "interposable_f", "noninterposable_f"], \
-; RUN:  "m2_f1":["m1_f1", "m1_f2"]}' > %t/workload_defs.json
-; RUN: llvm-lto2 run %t/workload1.bc %t/workload2.bc %t/workload3.bc \
-; RUN:  -thinlto-workload-def=%t/workload_defs.json -o %t/result.o -save-temps \
-; RUN:  -r %t/workload1.bc,m1_f1,plx \
-; RUN:  -r %t/workload1.bc,interposable_f \
-; RUN:  -r %t/workload1.bc,noninterposable_f \
-; RUN:  -r %t/workload1.bc,m1_variant \
-; RUN:  -r %t/workload2.bc,m2_f1,plx \
-; RUN:  -r %t/workload2.bc,m2_f1_alias \
-; RUN:  -r %t/workload2.bc,interposable_f,p \
-; RUN:  -r %t/workload2.bc,noninterposable_f,p \
-; RUN:  -r %t/workload2.bc,m2_variant \
-; RUN:  -r %t/workload3.bc,m1_f1 \
-; RUN:  -r %t/workload3.bc,m3_f1,plx
-; RUN: llvm-dis %t/result.o.1.3.import.bc -o - | FileCheck %s --check-prefix=FIRST
-; RUN: llvm-dis %t/result.o.2.3.import.bc -o - | FileCheck %s --check-prefix=SECOND
-; RUN: llvm-dis %t/result.o.3.3.import.bc -o - | FileCheck %s --check-prefix=THIRD
-;
-; The third module is bitwse-identical to the "normal" run, as the 
-; RUN: diff %t_baseline/result.o.3.3.import.bc %t/result.o.3.3.import.bc
-;
-; This time, we expect m1 to have m2_f1 and the m2 variant of interposable_f,
-; while keeping its variant of noninterposable_f
-;
-; FIRST-LABEL: @m1_f1
-; FIRST-LABEL: @m1_f2
-; FIRST-LABEL: define available_externally void @noninterposable_f
-; FIRST-NEXT: call void @m1_variant
-; FIRST-LABEL: @m2_f1
-; FIRST-LABEL: define available_externally void @interposable_f
-; FIRST-NEXT: call void @m2_variant
-; FIRST-NOT:   @m2_f1_alias
+; RUN: echo '{ \
+; RUN:    "m1_f1": ["m1_f1", "m2_f1", "m2_f1_alias", "interposable_f", "noninterposable_f"], \
+; RUN:    "m2_f1": ["m1_f1", "m1_f2", "interposable_f"] \
+; RUN:  }' > %t_exp/workload_defs.json
+;
+; RUN: llvm-lto2 run %t/m1.bc %t/m2.bc %t/m3.bc \
+; RUN:  -o %t_exp/result.o -save-temps \
+; RUN:  -thinlto-workload-def=%t_exp/workload_defs.json \
+; RUN:  -r %t/m1.bc,m1_f1,plx \
+; RUN:  -r %t/m1.bc,interposable_f,p \
+; RUN:  -r %t/m1.bc,noninterposable_f \
+; RUN:  -r %t/m1.bc,m1_variant \
+; RUN:  -r %t/m1.bc,m2_f1_alias \
+; RUN:  -r %t/m2.bc,m2_f1,plx \
+; RUN:  -r %t/m2.bc,m2_f1_alias,plx \
+; RUN:  -r %t/m2.bc,interposable_f \
+; RUN:  -r %t/m2.bc,noninterposable_f,p \
+; RUN:  -r %t/m2.bc,m2_variant \
+; RUN:  -r %t/m3.bc,m1_f1 \
+; RUN:  -r %t/m3.bc,m3_f1,plx
+; RUN: llvm-dis %t_exp/result.o.1.3.import.bc -o - | FileCheck %s --check-prefix=FIRST
+; RUN: llvm-dis %t_exp/result.o.2.3.import.bc -o - | FileCheck %s --check-prefix=SECOND
+; RUN: llvm-dis %t_exp/result.o.3.3.import.bc -o - | FileCheck %s --check-prefix=THIRD
+;
+; The third module is bitwse-identical to the "normal" run, as the workload
+; defintion doesn't mention it.
+;
+; RUN: diff %t_baseline/result.o.3.3.import.bc %t_exp/result.o.3.3.import.bc
+;
+; This time, we expect m1 to have m2_f1 and the m2 variant of both interposable_f
+; and noninterposable_f
+;
+; FIRST-LABEL:  @m1_f1
+; FIRST-LABEL:  @m1_f2.llvm.0
+;
+; @interposable_f is prevailing in m1, so it won't be imported
+; FIRST-LABEL:  define void @interposable_f
+; FIRST-NEXT:   call void @m1_variant
+;
+; FIRST-LABEL:  @m2_f1
+;
+; @noninterposable_f is prevailing in m2 so it will be imported from there. 
+; FIRST-LABEL:  define available_externally void @noninterposable_f
+; FIRST-NEXT:   call void @m2_variant
+;
+; FIRST-LABEL:  define available_externally void @m2_f1_alias
 ; SECOND-LABEL: @m2_f1
----------------
teresajohnson wrote:

Add comment for what we expect in the SECOND case. Ditto below for THIRD. Analogous to your "This time, we expect m1 ..." comment for FIRST above.

https://github.com/llvm/llvm-project/pull/74545