On Tue, 27 May 2014, Michael Schmitz wrote:
Hi Finn, according to the old driver source, all Ataris except for the TT use a 3.672 MHz clock at RTxC[A,B] for both the 9600 and 38400 baud settings. Most lower baud rates use the 8 MHz PCLK input. Divisors are 24 (9600 baud) and 6 (38400 baud) in that case.
Not quite (see below).
The TT uses that same arrangement on channel A, but something very different on channel B - 307.2 kHz at RTxCB and 2.4576 MHz on TRxCB. Divisors are 2 (using RTxCB) and 4 (using TRxCB), resepectively.
TT aside, is it safe to assume in head.S that channel B is clocked the same as channel A (i.e. 8 MHz)?
Adding all that in would complicate the code too much - let's settle for a comment in the code to say SCC_B should not be used on the TT without fixing the init table.
A comment is probably a good idea but it can be a different patch. That is, I'm happy for Geert to merge v2 unless you want to pursue 38400 baud.
1 file changed, 21 insertions(+), 10 deletions(-) Index: linux-m68k/arch/m68k/kernel/head.S =================================================================== --- linux-m68k.orig/arch/m68k/kernel/head.S 2014-05-26 23:13:50.000000000 +1000 +++ linux-m68k/arch/m68k/kernel/head.S 2014-05-26 23:24:33.000000000 +1000 @@ -2722,6 +2722,7 @@ func_return get_new_page #define MAC_USE_SCC_B /* Printer port */ #if defined(MAC_USE_SCC_A) || defined(MAC_USE_SCC_B) +/* Initialisation table for SCC with 3.6864 MHz PCLK */ L(scc_initable_mac): .byte 4,0x44 /* x16, 1 stopbit, no parity */ .byte 3,0xc0 /* receiver: 8 bpc */ @@ -2744,14 +2745,12 @@ L(scc_initable_mac): #define USE_MFP #if defined(USE_SCC_A) || defined(USE_SCC_B) -#define USE_SCC -/* Initialisation table for SCC */ -L(scc_initable): - .byte 9,12 /* Reset */ +/* Initialisation table for SCC with 7.9872 MHz PCLK */ +/* PCLK == 8.0539 gives baud == 9680.1 */ +L(scc_initable_atari): .byte 4,0x44 /* x16, 1 stopbit, no parity */ .byte 3,0xc0 /* receiver: 8 bpc */ .byte 5,0xe2 /* transmitter: 8 bpc, assert dtr/rts */ - .byte 9,0 /* no interrupts */ .byte 10,0 /* NRZ */ .byte 11,0x50 /* use baud rate generator */ .byte 12,24,13,0 /* 9600 baud */Would be 6 instead of 24 here for 38400, presumably?
From http://linux-mac68k.cvs.sourceforge.net/viewvc/linux-mac68k/linux-mac68k/drivers/char/atari_SCC.c
351 For 8.053976 MHz == base 503374: 352 0 bps -> 0 353 50 bps -> 5032 354 75 bps -> 3354 355 110 bps -> 2286 356 134 bps -> 1876 357 150 bps -> 1676 358 200 bps -> 1256 359 300 bps -> 837 360 600 bps -> 417 361 1200 bps -> 208 362 1800 bps -> 138 363 2400 bps -> 103 364 4800 bps -> 50 365 9600 bps -> 24 366 19200 bps -> 11 367 31500 bps -> 6 (really 31461 bps) 368 50000 bps -> 3 369 125000 bps -> 0 Based on Baud = PCLK / (2 x mode x (Time Constant + 2)), we have: mode 16 16 16 16 16 16 TC 6 11 24 1 4 10 PCLK 8.0539 8.0539 8.0539 3.670 3.670 3.670 MHz baud 31460.5 19360.3 9680.2 38250.0 19125.0 9562.5 The x16 mode can't get close to 38400 baud. The first version of this patch used the x1 mode: mode 1 1 1 TC 102 102 103 PCLK 7.9872 8.0539 8.0539 MHz baud 38400.0 38720.7 38351.9 I don't know why 38720.7 wasn't close enough to work when apparently 9680.2 baud does work. Anyway, TC = 103 should be closer to 38400 baud given we now know PCLK is 8.0539 MHz, unless I'm missing something... -- -- To unsubscribe from this list: send the line "unsubscribe linux-m68k" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
