updated das16 and cb-pcidas info

diff --git a/Documentation/comedi/drivers.txt b/Documentation/comedi/drivers.txt
index 8f22f213fce5b665a4ab863629ae98c24a2ce398..40520f9e492b4800772db13dce9a581f9d8ee5d6 100644 (file)
--- a/Documentation/comedi/drivers.txt
+++ b/Documentation/comedi/drivers.txt
@@ -84,12 +84,11 @@ Configuration options:
 
 Driver: cb_pcidas.o
 Description: ComputerBoards PCI-DAS1200, PCI-DAS1200/JR, may work with other PCI-DAS
-Author: Ivan Martinez <ivanmr@altavista.com>
-Status: works with PCI-DAS1200
-       should work with PCI-DAS1200/JR, not tested
-       may work with other PCI-DAS
+Author: Ivan Martinez <ivanmr@altavista.com>, Frank Mori Hess <fmhess@uiuc.edu>
+Status: Works with PCI-DAS1200.
+       Should work with PCI-DAS1200/JR, and PCI-DAS100X but not tested.
+       May work with other PCI-DAS-XXX.
 
-Doesn't support commands.
 This intends to be the driver for the whole PCI-DAS series. However,
 it was programmed with the manuals for PCI-DAS1200 and PCI-DAS1200/JR,
 and tested with the first one. IF YOU TRY THE DRIVER WITH ANOTHER PCI-DAS
@@ -240,10 +239,17 @@ Status: unknown
 
 Driver: das16.o
 Description: DAS16 compatible boards
-Authors: Sam Moore, Warren Jasper, ds, Chris Baugher
+Authors: Sam Moore, Warren Jasper, ds, Chris Baugher, Frank Hess
 Status: in development
 
 A rewrite of the das16 and das1600 drivers.
+Options:
+       [0] - base io address
+       [1] - irq (optional)
+       [2] - dma (optional)
+       [3] - master clock speed in MHz (1 or 10, defaults to 1)
+Both an irq line and dma channel are required for timed or externally
+triggered conversions.
 
 
 
@@ -257,12 +263,15 @@ As far as I know, there are no other boards that have
 the same register layout.  Even the CIO-DAS16/M1/16 is
 significantly different.
 
-Don't expect to reach the full 1 MHz capability of this
-board.  It can't do dma, so at 1 MHz your cpu is going
-to be spending almost all of its time trying to pull
-samples across the ISA bus.  I have gotten it to run
-at 833kHz though, using a hard real-time interrupt
-(TRIG_RT flag with rtlinux).
+I was _barely_ able to reach the full 1 MHz capability
+of this board, using a hard real-time interrupt
+(set the TRIG_RT flag in your comedi_cmd and use
+rtlinux or RTAI).  The board can't do dma, so the bottleneck is
+pulling the data across the ISA bus.  I timed the interrupt
+handler, and it took my computer ~470 microseconds to pull 512
+samples from the board.  So at 1 Mhz sampling rate,
+expect your CPU to be spending almost all of its
+time in the interrupt handler.
 
 Options:
        [0] - base io address