[Bug translator/34153] New: speed up tracepoint-based syscalls

[fche at redhat dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=34153

            Bug ID: 34153
           Summary: speed up tracepoint-based syscalls
           Product: systemtap
           Version: unspecified
            Status: NEW
          Severity: normal
          Priority: P2
         Component: translator
          Assignee: systemtap at sourceware dot org
          Reporter: fche at redhat dot com
  Target Milestone: ---

At this point, tp_syscalls are implemented as independent handlers for
kernel.trace("sys_enter") apprx., starting with a "gate" predicate that rejects
the handler if the current syscall# is not the desired one.  For tracing one
syscall that's ok, but for tracing hundreds, that's an O(N) slowdown via linear
search for a handler that will pass.

Seems like we need an "if-chain to switch conversion"-equivalent optimization
here.

-- 
You are receiving this mail because:
You are the assignee for the bug.