From: Pedro Alves Date: Mon, 23 Feb 2015 18:59:38 +0000 (+0000) Subject: gdbserver: redo stepping over breakpoint that was on top of a permanent breakpoint X-Git-Url: https://git.libre-soc.org/?a=commitdiff_plain;h=8090aef2bf5021f35c94193a035eb1ecd5e25e41;p=binutils-gdb.git gdbserver: redo stepping over breakpoint that was on top of a permanent breakpoint I'm going to add an alternate mechanism of breakpoint trap identification to 'check_stopped_by_breakpoint' that does not rely on checking the instruction at PC. The mechanism currently used to tell whether we're stepping over a permanent breakpoint doesn't fit in that new method. This patch redoes the whole logic in a different way that works with both old and new methods, in essence moving the "stepped permanent breakpoint" detection "one level up". It makes lower level check_stopped_by_breakpoint always the adjust the PC, and then has linux_wait_1 advance the PC past the breakpoint if necessary. This ends up being better also because this now handles non-decr_pc_after_break targets too. Before, such targets would get stuck forever reexecuting the breakpoint instruction. Tested on x86_64 Fedora 20. gdb/gdbserver/ChangeLog: 2015-02-23 Pedro Alves * linux-low.c (check_stopped_by_breakpoint): Don't check if the thread was doing a step-over; always adjust the PC if we stepped over a permanent breakpoint. (linux_wait_1): If we stepped over breakpoint that was on top of a permanent breakpoint, manually advance the PC past it. --- diff --git a/gdb/gdbserver/ChangeLog b/gdb/gdbserver/ChangeLog index d724e6c8c25..a130aab503c 100644 --- a/gdb/gdbserver/ChangeLog +++ b/gdb/gdbserver/ChangeLog @@ -1,3 +1,11 @@ +2015-02-23 Pedro Alves + + * linux-low.c (check_stopped_by_breakpoint): Don't check if the + thread was doing a step-over; always adjust the PC if + we stepped over a permanent breakpoint. + (linux_wait_1): If we stepped over breakpoint that was on top of a + permanent breakpoint, manually advance the PC past it. + 2015-02-23 Pedro Alves * linux-x86-low.c (REGSIZE): Define in both 32-bit and 64-bit diff --git a/gdb/gdbserver/linux-low.c b/gdb/gdbserver/linux-low.c index a6e1e3dce5f..1c66985a11f 100644 --- a/gdb/gdbserver/linux-low.c +++ b/gdb/gdbserver/linux-low.c @@ -507,14 +507,8 @@ check_stopped_by_breakpoint (struct lwp_info *lwp) /* We may have just stepped a breakpoint instruction. E.g., in non-stop mode, GDB first tells the thread A to step a range, and then the user inserts a breakpoint inside the range. In that - case, we need to report the breakpoint PC. But, when we're - trying to step past one of our own breakpoints, that happens to - have been placed on top of a permanent breakpoint instruction, we - shouldn't adjust the PC, otherwise the program would keep - trapping the permanent breakpoint forever. */ - if ((!lwp->stepping - || (!ptid_equal (ptid_of (current_thread), step_over_bkpt) - && lwp->stop_pc == sw_breakpoint_pc)) + case we need to report the breakpoint PC. */ + if ((!lwp->stepping || lwp->stop_pc == sw_breakpoint_pc) && (*the_low_target.breakpoint_at) (sw_breakpoint_pc)) { if (debug_threads) @@ -2552,6 +2546,41 @@ linux_wait_1 (ptid_t ptid, return ptid_of (current_thread); } + /* If step-over executes a breakpoint instruction, it means a + gdb/gdbserver breakpoint had been planted on top of a permanent + breakpoint. The PC has been adjusted by + check_stopped_by_breakpoint to point at the breakpoint address. + Advance the PC manually past the breakpoint, otherwise the + program would keep trapping the permanent breakpoint forever. */ + if (!ptid_equal (step_over_bkpt, null_ptid) + && event_child->stop_reason == LWP_STOPPED_BY_SW_BREAKPOINT) + { + unsigned int increment_pc; + + if (the_low_target.breakpoint_len > the_low_target.decr_pc_after_break) + increment_pc = the_low_target.breakpoint_len; + else + increment_pc = the_low_target.decr_pc_after_break; + + if (debug_threads) + { + debug_printf ("step-over for %s executed software breakpoint\n", + target_pid_to_str (ptid_of (current_thread))); + } + + if (increment_pc != 0) + { + struct regcache *regcache + = get_thread_regcache (current_thread, 1); + + event_child->stop_pc += increment_pc; + (*the_low_target.set_pc) (regcache, event_child->stop_pc); + + if (!(*the_low_target.breakpoint_at) (event_child->stop_pc)) + event_child->stop_reason = LWP_STOPPED_BY_NO_REASON; + } + } + /* If this event was not handled before, and is not a SIGTRAP, we report it. SIGILL and SIGSEGV are also treated as traps in case a breakpoint is inserted at the current PC. If this target does