1181834SrobertoProblem with recent ANSI compilers 2181834Sroberto 3181834SrobertoOn some systems, including AIX, the compiler quits on the ntp_refclock.c 4181834Srobertofile when processing the refclock_report() routine. The problem, which 5181834Srobertois eithre a feature or a bug, has to do with an unwanted promotion of 6181834Srobertothe u_char argument to an int and a failure of the compiler to recognize 7181834Srobertothe preceding prototype. A workaround is to use ANSI syntax to delare 8181834Srobertothe arguments. Since ANSI compilers are not universally available, this 9181834Srobertosyntax can't be used in the stock distribution. 10181834Sroberto 11181834Sroberto(Message # 60: 2884 bytes, New) 12181834SrobertoDate: Sat, 19 Aug 1995 13:20:50 -0400 13181834SrobertoFrom: "R. Bernstein" <rocky@panix.com> 14181834SrobertoNewsgroups: comp.protocols.time.ntp 15181834Srobertoto: mills@udel.edu 16181834Srobertoreturn-receipt-to: rocky@panix.com 17181834SrobertoSubject: time and AIX 3.2.5 raw tty bug 18181834Sroberto 19181834SrobertoThis posting isn't strictly about NTP, any program that may stop the 20181834Srobertoclock or set the clock backwards is subject to the AIX 3.2.5 bug. 21181834Sroberto 22181834SrobertoOn AIX 3.2.5, there is a bug in the tty driver for a raw device which 23181834Srobertomay crash the box under certain conditions: basically a read() on a 24181834Srobertoraw tty in effect, a character was read but not as many as specified 25181834Srobertoby VMIN when a read timeout occurred. VTIME specifies the timeout. See 26181834Srobertothe AIX manual page on termios.h or that include file. for Information 27181834Srobertoon VMIN (or MIN) VTIME (or TIME). 28181834Sroberto 29181834SrobertoA remedy other than to not use raw tty's is to apply patch U435110. 30181834Sroberto 31181834SrobertoDetails of the problem report follow. 32181834Sroberto 33181834Sroberto> ABSTRACT: 34181834Sroberto> IX43779: TRAP IN PSX_TIMEO 35181834Sroberto> 36181834Sroberto> ORIGINATING DETAILS: 37181834Sroberto> Stacktrace shows: 38181834Sroberto> IAR: 01460214 posixdd:psx_timeo + 8bf4: ti 4,r12,0x0 39181834Sroberto> *LR: 014601a0 posixdd:psx_timeo + 8b80 40181834Sroberto> 00212c60: 014604f4 posixdd:psx_timer + 8ed4 41181834Sroberto> 00212cc0: 0144b74c ttydd:tty_do_offlevel + 4284 42181834Sroberto> 00212d20: 000216fc .i_offlevel + 8c 43181834Sroberto> 00212d70: 00021d78 .i_softint + c8 44181834Sroberto> 00001004: 00008714 .finish_interrupt + 80 45181834Sroberto> 46181834Sroberto> RESPONDER SUMMARY: 47181834Sroberto> AIX asserted in psx_timeo(). Reason for the assert was that 48181834Sroberto> the current time was behind psx_ctime. Since this state 49181834Sroberto> can occur when the current time is changed after a character 50181834Sroberto> is received but before the VTIME interbyte timer pops, we 51181834Sroberto> should not assert on this. 52181834Sroberto> 53181834Sroberto> RESPONDER CONCLUSION: 54181834Sroberto> Removed the requirement that current time > psx_ctime by 55181834Sroberto> adding a new L_ntimersub macro that is used instead of the 56181834Sroberto> ntimersub macro in time.h. Also added a test for (current 57181834Sroberto> time - psx_ctime) being negative, in that case we do not 58181834Sroberto> adjust the new timeout. 59181834Sroberto> 60181834Sroberto> Reported to Correct a PTF in Error: NO 61181834Sroberto> Reported as a Highly pervasive problem: NO 62181834Sroberto> 63181834Sroberto> PE Apar?: NoPE 64181834Sroberto> Hiper Apar?: NoHiper 65181834Sroberto> Status: CLOSED PER 66181834Sroberto> Component Name: AIX V3 FOR RS/6 67181834Sroberto> Version: 320 68181834Sroberto> Component ID: 575603001 69181834Sroberto> Submitted: 94/05/03 70181834Sroberto> Closed: 94/05/05 71181834Sroberto> ChangeTeam: TX2527 72181834Sroberto> 73181834Sroberto> APAR FIXED BY: U431696 U432151 U432844 U432870 U432979 74181834Sroberto> U433049 U433081 U433459 U433876 U433906 U434598 U434453 75181834Sroberto> U434672 U434737 U435110 76181834Sroberto 77