Repatch (from linux) of OSX IRAF

1 files changed, 60 insertions, 0 deletions

diff --git a/sys/ki/kbzard.x b/sys/ki/kbzard.x
new file mode 100644
index 00000000..7370f7f9
--- /dev/null
+++ b/sys/ki/kbzard.x
@@ -0,0 +1,60 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<mach.h>
+include	<config.h>
+include	"ki.h"
+
+# KB_ZARD -- Asynchronous read from a binary device.  We are called only if
+# the device does not reside on the local node.  In this implementation
+# these reads are not actually asynchronous.  This is possible but more
+# complex to implement than I wanted to get into at this time, particularly
+# since FIO is not yet fully asynchronous.  The problem is that a node
+# channel may be used to multiplex any number of requests to the remote
+# node process.  If an asynchronous read is pending this must be detected
+# and the read completed before processing any other requests.
+
+procedure kb_zard (device, chan, obuf, max_bytes, loffset)
+
+int	device			# device driver code
+int	chan			# channel assigned device
+char	obuf[max_bytes]		# receives data
+int	max_bytes		# max bytes to read
+long	loffset			# file offset
+
+int	server, status
+int	ki_send(), ki_receive()
+include	"kichan.com"
+include	"kii.com"
+
+begin
+	server = k_node[chan]
+
+	if (max_bytes <= 0) {
+	    k_status[chan] = 0
+	    return
+	}
+
+	# Send the request to initiate the read.
+
+	p_arg[1] = k_oschan[chan]
+	p_arg[2] = max_bytes
+	p_arg[3] = loffset
+
+	if (ki_send (server, device, BF_ARD) == ERR) {
+	    status = ERR
+	} else {
+	    # Wait for the ZAWT packet.
+	    if (ki_receive (server, device, BF_AWT) == ERR)
+		status = ERR
+	    else
+		status = p_arg[1]
+
+	    # Read the data block (if any) directly into caller's buffer.
+	    if (status > 0) {
+		call ks_aread (server, obuf, status)
+		call ks_await (server, status)
+	    }
+	}
+
+	k_status[chan] = status
+end