[Bug target/82516] New: Optimize jmp chain

[antoshkka at gmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=82516

            Bug ID: 82516
           Summary: Optimize jmp chain
           Product: gcc
           Version: 8.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: target
          Assignee: unassigned at gcc dot gnu.org
          Reporter: antoshkka at gmail dot com
  Target Milestone: ---

Following code

unsigned bar();
unsigned foo1(int i) {
    if (i < 0) {
        return bar();
    }
    return i;
}

at -O2 produces the assembly:

foo1(int):
        test    edi, edi
        js      .L4     <== use "js bar()" instead
        mov     eax, edi
        ret
.L4:
        jmp     bar()   <== this could be removed


Generally speaking, any chain of jmps could be optimized to directly jumping to
a target. For example

js A
A: jmp B
B: jmp C

could be optimized to "js C"