maxjhandsome
/
llvm-project
forked from OSchip/llvm-project
1
0
Fork 0
Code Issues Pull Requests Packages Releases Wiki Activity
llvm-project/llvm/test/ThinLTO/X86/funcattrs-prop-weak.ll

Ignoring revisions in .git-blame-ignore-revs. Click here to bypass and see the normal blame view.

61 lines
2.1 KiB
LLVM
Raw Normal View History

[ThinLTO] Add noRecurse and noUnwind thinlink function attribute propagation Thinlink provides an opportunity to propagate function attributes across modules, enabling additional propagation opportunities. This change propagates (currently default off, turn on with `disable-thinlto-funcattrs=1`) noRecurse and noUnwind based off of function summaries of the prevailing functions in bottom-up call-graph order. Testing on clang self-build: 1. There's a 35-40% increase in noUnwind functions due to the additional propagation opportunities. 2. Throughput is measured at 10-15% increase in thinlink time which itself is 1.5% of E2E link time. Implementation-wise this adds the following summary function attributes: 1. noUnwind: function is noUnwind 2. mayThrow: function contains a non-call instruction that `Instruction::mayThrow` returns true on (e.g. windows SEH instructions) 3. hasUnknownCall: function contains calls that don't make it into the summary call-graph thus should not be propagated from (e.g. indirect for now, could add no-opt functions as well) Testing: Clang self-build passes and 2nd stage build passes check-all ninja check-all with newly added tests passing Reviewed By: tejohnson Differential Revision: https://reviews.llvm.org/D36850
2021-09-28 03:24:28 +08:00
; RUN: split-file %s %t
; RUN: opt -thinlto-bc %t/a.ll -thin-link-bitcode-file=%t1.thinlink.bc -o %t1.bc
; RUN: opt -thinlto-bc %t/b.ll -thin-link-bitcode-file=%t1.thinlink.bc -o %t2.bc
; RUN: opt -thinlto-bc %t/c.ll -thin-link-bitcode-file=%t1.thinlink.bc -o %t3.bc
; If the prevailing weak symbol is defined in a native file, the IR copies should be dead and propagation should not occur
; RUN: llvm-lto2 run -disable-thinlto-funcattrs=0 %t1.bc %t2.bc %t3.bc -o %t.o \
; RUN: -r %t1.bc,caller,px -r %t1.bc,callee,lx \
; RUN: -r %t2.bc,callee,x \
; RUN: -r %t3.bc,callee,x \
; RUN: -save-temps
; RUN: llvm-dis -o - %t.o.1.3.import.bc | FileCheck %s
; If the prevailing weak symbol is in an IR file, it should be the one used in the final binary and thus propagation
; should be based off of that copy
; RUN: llvm-lto2 run -O3 -disable-thinlto-funcattrs=0 %t1.bc %t2.bc %t3.bc -o %t.2.o \
; RUN: -r %t1.bc,caller,px -r %t1.bc,callee,lx \
; RUN: -r %t2.bc,callee,px \
; RUN: -r %t3.bc,callee,x \
; RUN: -save-temps
; RUN: llvm-dis -o - %t.2.o.1.3.import.bc | FileCheck %s --check-prefix=PREVAILING
; RUN: llvm-dis -o - %t.2.o.2.3.import.bc | FileCheck %s --check-prefix=PREVAILING-B
; RUN: llvm-dis -o - %t.2.o.3.3.import.bc | FileCheck %s --check-prefix=PREVAILING-C
;--- a.ll
target datalayout = "e-m:e-i64:64-f80:128-n8:16:32:64-S128"
target triple = "x86_64-unknown-linux-gnu"
declare i32 @callee()
; CHECK-NOT: Function Attrs:
; CHECK: define i32 @caller()
; PREVAILING: Function Attrs: norecurse nounwind
; PREVAILING-NEXT: define i32 @caller()
define i32 @caller() {
%res = call i32 @callee()
ret i32 %res
}
;--- b.ll
target datalayout = "e-m:e-i64:64-f80:128-n8:16:32:64-S128"
target triple = "x86_64-unknown-linux-gnu"
; PREVAILING-B: define weak i32 @callee()
define weak i32 @callee() {
ret i32 5
}
;--- c.ll
target datalayout = "e-m:e-i64:64-f80:128-n8:16:32:64-S128"
target triple = "x86_64-unknown-linux-gnu"
; PREVAILING-C: declare i32 @callee()
define weak i32 @callee() {
ret i32 6
}