Lines Matching full:resolved
50 * branch to resolved address. That means that on first hit, we must
63 * a curried call to the resolver. When the symbol is resolved, the
79 * and resolved (to some address). When the process starts (or when
83 * we assume it was already resolved, and convert the breakpoint to
84 * resolved. We also rewrite the resolved value in .plt back to the
87 * When a PLT entry hits a resolved breakpoint (which happens because
99 * resolved, and rewrite the .plt value back to PLT address. We then
103 * resolved, and put a breakpoint there. The next time around (when
104 * the next PLT entry is to be resolved), instead of single-stepping
106 * this breakpoint. When it hits, we know the PLT entry was resolved.
111 * (actually .iplt entry, these live in a special section) is resolved
115 * catching the point where these get resolved to unresolve them. So
120 * catch the point where the slot is resolved, would hit the return
162 * case some slots will have been resolved already.
169 * all the PLT slots are it). When resolved, it looks either
182 * is resolved or not. */
281 * ourselves with bias, as the values in OPD have been resolved
614 /* We only modify plt_entry[0], which holds the resolved
841 /* We mark the breakpoint as resolved on both arches. */
877 debug(DEBUG_PROCESS, "pid=%d PLT got resolved to value %#"PRIx64,
880 /* The .plt slot got resolved! We can migrate the breakpoint
881 * to RESOLVED and stop single-stepping. */
