Any objections to the patch at http://www.FreeBSD.org/~jhb/patches/set_dbregs.patch ? First, the suser() call in the existing code is bogus as it should be using curthread (who's doing the work) rather than td (who's watchpoint is being set). (Side note: we aren't doing anything to keep the debug registers in a multithreaded process in sync, so watch points for multithreaded apps in gdb is probably busted right now.) Secondly, why on earth would it really be a good idea for a userland process to set a hardware watch point in KVA even if they are root? Shouldn't this be the job of a kernel debugger interfacing to KDB (such as remote gdb)? -- John Baldwin <jhb_at_FreeBSD.org> <>< http://www.FreeBSD.org/~jhb/ "Power Users Use the Power to Serve" = http://www.FreeBSD.orgReceived on Tue Mar 07 2006 - 20:51:08 UTC
This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:38:53 UTC