This is the mail archive of the ecos-discuss@sourceware.cygnus.com mailing list for the eCos project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]

Re: Linux Synthetic target device support!


>>>>> "Mohammed" == Mohammed Illyas Mansoor <mansoor@cdotb.ernet.in> writes:
Mohammed>     Ok I found ser_filter.cpp it was in
Mohammed> <base>/tools/src/eCosTest dir made ser_filter and tested
Mohammed> serial3-5.

Mohammed>     serial3 is passing only if BINARY data is within 120
Mohammed> range ser_filter is timing out if data is more than 120.
Mohammed> <cut> [f] 0000 40 42 49 4e 41 52 59 3a '@BINARY:' [f] 0008
Mohammed> 31 32 38 3a 31 21 .. .. '128:1!'  [f] Dispatching command
Mohammed> BINARY.  [f] **** Timed out while reading -- resuming as
Mohammed> null filter [f] Data received 127 bytes (of 1) from target:
Mohammed> </cut>

Interesting. It may be because there's a buffer overflow, or a problem
with performance. Sorry, I've forgotten what hardware you are using -
but try to force the test to run at 9600 baud to see if it makes a
difference (see serial4.c for similar hacks for AEB).

Mohammed>     serial4 test was going in a loop trying to set_config
Mohammed> the serial device, I tried to figure out in serial4.c there
Mohammed> it was checking for R , S or D char's from ser_filter but
Mohammed> ser_filter was actually sending OK for change_config
Mohammed> command.

All protocol exchange actually happens in ser_test_protocol.inl. The
code there may help you track down the problem.

Cheers,
Jesper

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]