File: symbolize-split-dwarf-no-skel-address.test

package info (click to toggle)
swiftlang 6.1.3-2
  • links: PTS, VCS
  • area: main
  • in suites: forky, sid
  • size: 2,791,604 kB
  • sloc: cpp: 9,901,740; ansic: 2,201,431; asm: 1,091,827; python: 308,252; objc: 82,166; f90: 80,126; lisp: 38,358; pascal: 25,559; sh: 20,429; ml: 5,058; perl: 4,745; makefile: 4,484; awk: 3,535; javascript: 3,018; xml: 918; fortran: 664; cs: 573; ruby: 396
file content (22 lines) | stat: -rw-r--r-- 877 bytes parent folder | download | duplicates (26) 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
 
Copy the DWO and then cd into the output directory, so that the symbolizer would
find the DWO if it attempted to.
RUN: rm -rf %t && mkdir -p %t
RUN: cd %t
RUN: cp %p/Inputs/split-dwarf-no-skel-address.dwo %t
RUN: llvm-symbolizer --inlining --no-demangle \
RUN:    --obj=%p/Inputs/split-dwarf-no-skel-address.o 0x4 | FileCheck %s

Built from the following source:
void f1();
__attribute__((always_inline)) inline void f2() { f1(); }
void f3() { f2(); }

Compiled to assembly with clang, modified the skeleton CU to remove the
high/low pc (& update the CU length field and abbrev to match) & then
compile/objcopy to make the .o and .dwo.

Ensure that the f2 inlined frame is not included - it's inefficient to have to
go and load all the debug info and search for the address ranges, so assume
that a lack of ranges on the CU means the CU covers no addresses.

CHECK-NOT: _Z2f2v