[HECnet] Bizarre RSTS/E and SIMH issue...

Cory, you could try to dissable asynchronous IO (SET NOASYNCH) and see if the crash goes away. Anyways, I think this should be reported as a simh issue in the project github. Jordi Guillaumes i Pons Barcelona - Catalunya - Europa El 26/07/2013, a les 16:18, <Paul_Koning at Dell.com> va escriure: "tq" is a tape drive, I think, so the thing to do would be to find the tape emulation code where that message is being reported and stick some breakpoints into the spot where that happens. The message you're seeing is for the Unix error code EAGAIN, which normally shows up with non-blocking I/O -- if the operation would have resulted in blocking, you get EAGAIN instead if you asked for non-blocking mode. For example, a read when the data isn't available yet. This may be a case of setting non-blocking mode but not handing the resulting error returns correctly in all cases. paul On Jul 26, 2013, at 10:07 AM, Cory Smelosky wrote: On 26 Jul 2013, at 10:05, Paul_Koning <Paul_Koning at Dell.com<mailto:Paul_Koning at Dell.com>> wrote: If you dropped into the simh prompt, that would make it a simh bug. And that message looks like a Unix message; it doesn't look like a RSTS message. That s what I was thinking. Issue persists with both -O2 and -O0 so it s not GCC being stupid (for once). This is latest git issue isn t present in 3.9. paul On Jul 26, 2013, at 8:57 AM, Cory Smelosky wrote: Morning all, Restoring MAIL installation files from the MAIL Library kit TQ I/O error: Resource temporarily unavailable I/O error, PC: 102054 (SOB R1,102052) sim> I can't even begin to find what would cause this...is it a bug in SIMH or have I found the weirdest IPC error yet? Is it an underlying I/O error on the disk? -- Cory Smelosky http://gewt.net<http://gewt.net/> Personal stuff http://gimme-sympathy.org<http://gimme-sympathy.org/> Projects