2009-05-22 04:57:24 hardware flow problem of bf561
彬 张 (CHINA)
Message: 74447
Hi.
We meet some problems of the bf561 hardware flow. We use PF4 PF5 which are directly connected to MAX3232 to implement hardware flow .In the kernel , we have chosen the choice of UART0 hardware flow .But when we use a cross cable to connect the CDMA board and the DSP board, the modem doesn't work.
We can confirm that the modem is right ,the pppd script is right and PF4 PF5 are only connected to MAX3232. We want to konw if we can use PF4 PF5 directly connect to MAX3232?? Shall we do something before connet to MAX3232??
QuoteReplyEditDelete
2009-05-22 05:44:20 Re: hardware flow problem of bf561
Sonic Zhang (CHINA)
Message: 74448
Are you running Linux? Which release do you run?
QuoteReplyEditDelete
2009-05-22 07:02:58 Re: hardware flow problem of bf561
彬 张 (CHINA)
Message: 74450
Are you running Linux? Which release do you run?
--- Yes,we run it under uclinux ,the version is 2008R1. We use uart connect to MAX3232 and then use cross cable to communicate with CDMA modem. It works well. But when we use PF4 PF5 to implement CTS RTS hardwareflow. It doesn't work. I don't know why!!
QuoteReplyEditDelete
2009-05-22 13:15:16 Re: hardware flow problem of bf561
Robin Getz (UNITED STATES)
Message: 74454
Flow control (for those that are software controlled) really only works in PIO mode - are you using that, or DMA?
-Robin
QuoteReplyEditDelete
2009-05-22 20:40:40 Re: hardware flow problem of bf561
bin zhang (CHINA)
Message: 74466
Hi.
We just choose as follows.
< > 8250/16550 and compatible serial support
--- Non-8250 serial port support
<*> Blackfin serial port support
[ ] Console on Blackfin serial port
UART Mode (DMA mode) --->
--- Enable UART0
[*] Enable UART0 hardware flow control
(5) UART0 CTS pin
(4) UART0 RTS pin
< > Blackfin SPORT emulate UART (EXPERIMENTAL)
Shall we change the UART Mode to make a try????
TranslateQuoteReplyEditDelete
2009-05-23 18:41:59 Re: hardware flow problem of bf561
Robin Getz (UNITED STATES)
Message: 74481
Bin:
DMA will not work with flow control. If needs to be PIO.
-Robin
QuoteReplyEditDelete
2009-05-23 21:41:56 Re: hardware flow problem of bf561
bin zhang (CHINA)
Message: 74485
Hi.
I had just tried it,but it dosen't work also..........
TranslateQuoteReplyEditDelete
2009-05-24 23:54:02 Re: hardware flow problem of bf561
Sonic Zhang (CHINA)
Message: 74513
You can run rtscts_test and check the voltage of PF4 and PF5 by a voltage meter.
If the voltage is incorrect according to you setting, please check the GPIO configuration.
QuoteReplyEditDelete
2009-05-26 05:04:13 Re: hardware flow problem of bf561
bin zhang (CHINA)
Message: 74597
Hi.
We use rtscts_test to test the voltage of PF4 (cts) and PF5 (rts),the voltage of CTS is 3.3V ,and the RTS is 0.2V whatever you disable or enable it .We have checked the GPIO configuration,it's ringht!
So we think that the hardflow still doesn't work . Could you please give me any more advice?
TranslateQuoteReplyEditDelete
2009-05-26 05:35:34 Re: hardware flow problem of bf561
Sonic Zhang (CHINA)
Message: 74600
Then you need to check gpio configuration why the RTS is not high after your enabled it.
QuoteReplyEditDelete
2009-05-26 06:19:35 Re: hardware flow problem of bf561
bin zhang (CHINA)
Message: 74601
Sonic Zhang:
We have checked it for many times. Because we use uart of BF561 and PF4 PF5 to connect to the MAX3232, the circuit is very very simple ! We also confirm that PF4 PF5 are only be used as flow control pin ,they don't connect to any other units.
When we configure the kernel ,we enable the UART0 hardflow choice and choose PIO mode ,but ithe flow control doesn't work .the voltage of PF4 PF5 don't change when the power is on.
So what do you mean about gpio configuration besides the above??
TranslateQuoteReplyEditDelete
2009-05-26 06:28:34 Re: hardware flow problem of bf561
Sonic Zhang (CHINA)
Message: 74603
You may update the bfin_5xx driver to the SVN 2008R1 branch. There several bug fixing after 2008R1-RC8 release. Please check if that help. Otherwise, debug the bfin_5xx driver by kgdb or gdbproxy or printk. We don't have your hardware.
QuoteReplyEditDelete
2009-05-26 07:52:05 Re: hardware flow problem of bf561
Robin Getz (UNITED STATES)
Message: 74605
Bin:
Have a look in the MMRs after things have booted.
https://docs.blackfin.uclinux.org/doku.php?id=linux-kernel:debug
QuoteReplyEditDelete
2009-05-26 08:38:24 Re: hardware flow problem of bf561
bin zhang (CHINA)
Message: 74609
Robin :
I just take a look at the FIO0 registers.
FIO0_FLAG_C 0xc030 FIO0_INEN 0xc0b0 FIO0_MASKA_C 0x0080
FIO0_FLAG_D 0xc030 FIO0_MASKA_D 0x0080
FIO0_FLAG_S 0xc030 FIO0_MASKA_S 0x0080
FIO0_FLAG_T 0xc030 FIO0_MASKA_T 0x0080
others 0x0000
TranslateQuoteReplyEditDelete
2009-05-26 09:29:03 Re: hardware flow problem of bf561
bin zhang (CHINA)
Message: 74612
Hi
I do the test again . (update bfin_5xx.c and bfin_serial_5xx.h from /branch/2008R1 )
when I use root:>rtscts_test ttyBF0 -s
it shows RTS(disable) CTS(enable)
FIO0_FLAG_C 0xc020
FIO0_FLAG_D 0xc020
FIO0_FLAG_S 0xc020
FIO0_FLAG_T 0xc020
and then I use the command root:>rtscts_test ttyBF0 -e
it shows RTS(enable) CTS(enable)
FIO0_FLAG_C 0xc000
FIO0_FLAG_D 0xc000
FIO0_FLAG_S 0xc000
FIO0_FLAG_T 0xc000
The voltage of CTS is 3.3V and RTS is 0.2V . They don't change either during the test.