On Thu, 6 Dec 2001, Mike McCormack wrote: > > Hi All, > > yeah, looks like there's a bit of confusion between RTS_CONTROL_ENABLE > and RTS_CONTROL_HANDSHAKE in dlls/kernel/comm.c > > RTS_CONTROL_ENABLE does not mean we should use flow control, it means > we should turn on RTS and leave it on. Similarly with > DTR_CONTROL_ENABLE. Man, what confusing names. > > Is the (untested) patch attached helpful? > > Looks like there will be more issues with this... > > Mike > Yep. That is all very well for my 32 bit app, but the 16 bit - works OK, unless you try a comm trace, then it crashes, with this little beauty: trace:comm:WriteComm16 cid 3, ptr 0x40f7d21e, length 1487 wine_dbg_vprintf: debugstr buffer overflow (contents: 'trace:comm:WriteComm16 1A2845F7: 4:PUTDATA 541 AD_STATS 0.AS I guess 1407 bytes plus the trace message is more than we need. Also, the app shuts off CRTSCTS. I think there is still something wrong in [GS]etCommState16 and/or DCBtoDCB16. Lawson Probable user head space error. - Dennis A. Moore
