Subversion Repositories or1k

2003-02-24  Jonathan Larmour  <jifl@eCosCentric.com>

        * cdl/openbsd_net.cdl: Improve doc links.

2003-01-30  Jonathan Larmour  <jifl@eCosCentric.com>

        * src/sys/net/if_bridge.c (bridge_input): Learn source host addr,
        not dest host addr.

2002-12-14  Nick Garnett  <nickg@ecoscentric.com>

        * src/sys/netinet/tcp_usrreq.c (tcp_usrreq): Rearranged ifndef on
        PRU_SENSE case to keep the case in the switch and let it return a
        sensible result. Otherwise a stat() on a socket causes a panic().

2002-12-03  Gary Thomas  <gthomas@ecoscentric.com>

        * src/ecos/support.c (show_network_tables): New function used
        to print network information (interface, routing).

2002-05-21  Jesper Skov  <jskov@redhat.com>

        * src/lib/recv.c: Fix warning. socket.h declarations are not
        available to the kernel code. Added missing argument in the
        recvfrom call.

2002-05-14  Jesper Skov  <jskov@redhat.com>

        * include/netinet6/in6.h: Fixed warnings.

        * src/sys/kern/uipc_socket2.c (sblock): Fixed warning.
        * src/sys/netinet/tcp_input.c: Same.
        * src/sys/net/if_loop.c (looutput): Same.
        * src/sys/kern/sockio.c: Same.

        * include/sys/param.h: Include machine headers after the namespace
        changing macros to get the desired declarations, thus fixing
        compiler warnings.

2002-04-26  Gary Thomas  <gthomas@redhat.com>

        * src/sys/netinet/ip_input.c: Proper "extern" definition of 'ipstat'.

        * src/sys/netinet/raw_ip.c: 
        * src/sys/netinet/ip_output.c: 
        * include/netinet/ip_var.h: Use of ramdom IP packet id is optional.

2002-04-22  Gary Thomas  <gthomas@redhat.com>

        * src/sys/netinet/tcp_usrreq.c: 
        * src/sys/netinet/tcp_output.c: 
        * src/sys/netinet/tcp_input.c: 
        * src/sys/netinet/tcp_debug.c: 
        * include/netinet/tcp_debug.h: Fully conditionalize on TCPDEBUG.

2002-03-27  Jonathan Larmour  <jlarmour@redhat.com>

        * doc/openbsd.sgml: Manpages now live in "common" BSD package.

2002-03-19  Gary Thomas  <gthomas@redhat.com>

        * src/sys/net/route.c (cyg_route_reinit): Renamed from route_reinit.

2002-03-11  Hugo Tyson  <hmt@redhat.com> 

        [Case 107110]

        * src/sys/netinet/in.c (in_control): SIOCSIFADDR switch entry
        moved to before the scan for this same address being in the list
        already, along with Add and Delete (SIOCAIFADDR,SIOCDIFADDR) arms.
        It falls through into the same alloc-if-needed code anyway.
        Thus repeatedly setting the same address does not leak store.

        * src/sys/net/route.c (route_reinit): Rewrite to delete all routes
        individually rather than en-masse (leaking store).
        (rt_reinit_rtdelete): New function; callback for individual
        deletion.
        (rtioctl): Do not pass in a "route **" to return a pointer to the
        route removed or added; this results in an extra reference, by the
        returned pointer, and so a storeleak.
        (rtrequest): RTM_DELETE arm: do not free a gateway route if the
        gateway pointer is the same as the route itself - it gets freed
        *again* at the end of the routine if you do.  Just dec the refcnt.

2002-03-07  Gary Thomas  <gthomas@redhat.com>

        * include/sys/param.h: Privatize route_reinit(), arc4random().

2002-02-28  Hugo Tyson  <hmt@redhat.com>
2002-02-28  Martin Buck <martin.buck@ascom.ch>

        * src/sys/net/if_ethersubr.c (ether_input): [Bugzilla 60318]
        ether_input() schedules processing by calling schednetisr(FOO) and
        then queues the frame afterwards.  Where we cannot be preempted
        eg. in a DSR or in the original BSD kernel, this is OK.  In eCos
        the call to schednetisr(FOO) might cause a context switch *before*
        the frame is enqueued.  This "jams" one packet in the queue,
        delaying it until the next packet, and so on.

        This change makes ether_input() store up the schednetisr() arg
        until the end, and do the call then, much like it does with the
        queue to use.  Thanks Martin!

2002-02-22  Hugo Tyson  <hmt@redhat.com>

        * doc/openbsd.sgml: New file; separated stack-specific docs from
        monolithic net doc, and tidied up in various ways.

2002-02-15  Gary Thomas  <gthomas@redhat.com>

        * cdl/openbsd_net.cdl: Slight changes in layout required by split
        of network stacks.

2002-02-13  Jonathan Larmour  <jlarmour@redhat.com>
2002-02-13  Andrew Lunn  <andrew.lunn@ascom.ch>

        * include/sys/mbuf.h: Declare variables as extern in header.
        * src/sys/kern/uipc_mbuf.c: Define mbtypes[] and init it.

2002-02-05  Hugo Tyson  <hmt@redhat.com>

        * cdl/openbsd_net.cdl: Place this package below  CYGPKG_NET
        ie. the common networking package.  Only affects the view in the
        GUI CT, no big deal.

2002-01-28  Gary Thomas  <gthomas@redhat.com>

        * src/sys/netinet/ip_input.c: inet_ntoa() is now common.

        * cdl/openbsd_net.cdl: Renamed from net.cdl.  This package
        is now a proper sub-package, implementing a network stack.

2002-01-28  Jonathan Larmour  <jlarmour@redhat.com>

        * cdl/net.cdl: Clarify CYGPKG_NET_ROUTING description.

2002-01-08  Jonathan Larmour  <jlarmour@redhat.com>

        * tests/server_test.c (server_test): Correct port conversion.

2001-12-14  Hugo Tyson  <hmt@redhat.com>

        * src/sys/net/if_ethersubr.c (ether_output): Never ever print
        anything!  It can recurse unexpectedly if debugging over net.

2001-12-10  Andrew Lunn <andrew.lunn@ascom.ch>

        * cdl/net.cdl: Add configury to eliminate network timing stats.
        * src/include/machine/param.h: Timing statistics for things like 
        memcpy, mbuf_alloc etc now use configury to determine if the
        statistics code should be compiled.
        * src/ecos/support.c: Ditto.
        * tests/nc_test_slave.c: Only print timing stats when they
        are enabled. 
        * tests/tcp_echo.c: Ditto.
        * tests/nc_test_master.c: Allow building with SNMP lib which has
        its own gettimeofday(). This should be fixed better a different
        way but will do for now.
        
2001-12-06  Hugo Tyson  <hmt@redhat.com>

        * src/lib/bootp_support.c (init_net): Just re-tabbed and checked
        the addition to initialize DNS from DHCP info - it had already
        snuck into the file in the previous change from Andrew.  Include
        of netdb.h added also, from the patch in Bug 57019.  Also check
        CYGINT_ISO_DNS for the setup code instead of CYGPKG_ISO_DNS.

2001-12-04  Jonathan Larmour  <jlarmour@redhat.com>

        * src/lib/bootp_support.c: Check CYGINT_ISO_DNS (from
        <pkgconf/isoinfra.h>) instead of CYGPKG_ISO_DNS.

2001-12-03   Andrew Lunn  <Andrew.Lunn@ascom.ch>

        * src/lib/bootp_support.c: get_bootp_option(): Take the
        max length of data we want out of the bootp record.
        * src/lib/dhcp_prot.c: Pass the length we expect for an option.

2001-08-16  Anssi Pulkkinen <Anssi.Pulkkinen@ascom.ch>

        * src/lib/dhcp_prot.c: Make sure we use the newly calculated xid.
        
2001-08-02  Anand Srivastava <Anand.Srivastava@ascom.ch>

        * src/lib/dhcp_prot.c do_dhcp: ID calculation must be done only
        once. moved it out the loop.    
        
2001-11-21  Hugo Tyson  <hmt@redhat.com>

        * src/lib/dhcp_support.c (dhcp_mgt_entry): If we have an
        SNMPAGENT, must recycle it whenever we reinitialize all
        interfaces; call SnmpdShutDown() to cause this.
        Also re-initialize all loopback interfaces here too.    

2001-11-30  Hugo Tyson  <hmt@redhat.com>

        * src/lib/dhcp_prot.c (set_default_dhcp_tags): New routine to
        insert all the extra woffle we need neatly each time, keeping the
        state machine's case arms to the real work.  The new thing this
        sets is TAG_DHCP_PARM_REQ_LIST, with a list that matches the
        default set we get from a LINUX dhcpd.  This is because M$ servers
        need explicit requests for eg. TAG_GATEWAY.  Note that I have
        included configuration to override the set of tags requested here
        in the source, but not yet backed up by CDL - we'll see whether
        anyone needs this.
        (set_variable_tag): New routine to insert a variable pointed-to
        data item rather than an int of 1,2 or 4 bytes.
        (do_dhcp): Call set_default_dhcp_tags() every time we send a
        packet, rather than ad hoc additions in each state.

2001-11-29  Jonathan Larmour  <jlarmour@redhat.com>

        * include/machine/ansi.h: No longer require BSD string function
        compatibility macros (in fact they confuse things).

        * cdl/net.cdl: Require BSD compatibility functions from <string.h>.

2001-11-19  Hugo Tyson  <hmt@redhat.com>

        * include/bootp.h (BP_STD_TX_MINPKTSZ): New symbol defining the
        minimal DHCP packet size which we should send.

        * src/lib/dhcp_prot.c (scan_dhcp_size): New routine to scan up to
        the end of the packet, for length determination and padding.
        (dhcp_size): Now uses scan_dhcp_size().
        (dhcp_size_for_send): New, uses scan_dhcp_size() and pads with
        zero up to the min packet size that we should send.
        (do_dhcp): In every sendto() call, use dhcp_size_for_send(xmit) in
        order to send padded, full size packets as needed.  5 instances.

        Credit to "Anssi Pulkkinen" <Anssi.Pulkkinen@ascom.ch> for the
        original version of this patch.

2001-11-08  Jesper Skov  <jskov@redhat.com>

        * tests/server_test.c (net_test): Fix compile error.

2001-11-02  Gary Thomas  <gthomas@redhat.com>

        * src/ecos/support.c: 
        * include/machine/ansi.h: Add C++ support [externC].

2001-11-11  Andrew Lunn <andrew.lunn@ascom.ch>

        *src/lib/recv.c: Implement the recv() call.

2001-10-29  Hugo Tyson  <hmt@redhat.com>

        * src/sys/net/if_ethersubr.c (ether_output): [CASE 106613] Even if
        the queue is full, and we are dropping the packet, try to start
        the interface anyway, to give it a chance to empty the queue if
        the device has recovered from whatever made the queue fill up in
        the first place - being unplugged from the network for example.
        This change is belt & braces with a similar policy in the periodic
        tickle function in the logical ether driver in io/eth/... ; this
        change will recover the situation immediately if the application
        continues trying to send despite ENOBUFS.

2001-10-25  Hugo Tyson  <hmt@redhat.com>

        * src/sys/net/if_bridge.c (bridge_broadcast): Count if_obytes in
        destination interfaces as well as the bridge interface; normally
        this is done in if_ethersubr.c but here we talk directly to the
        interface, bypassing that layer.  Thanks to Andrew Lunn for
        noticing that.

2001-10-18  Jonathan Larmour  <jlarmour@redhat.com>

        * src/sys/netinet/in.c (in_lifaddr_ioctl): Silence warnings about
        trigraphs.

2001-10-10  Hugo Tyson  <hmt@redhat.com>

        * tests/ftp_test.c (net_test): This was lacking #ifdefs for
        CYGHWR_NET_DRIVER_ETH0 and CYGHWR_NET_DRIVER_ETH1.  Ooops.

2001-10-05  Jonathan Larmour  <jlarmour@redhat.com>

        * cdl/net.cdl (CYGPKG_NET_CFLAGS_ADD): Define __INSIDE_NET within
        this package to allow segregation of definitions only relevant for
        internal consumption, and definitions allowable for external
        consumption.

        * include/lib/libkern/libkern.h: Conditionalize on __INSIDE_NET.

2001-10-04  Jonathan Larmour  <jlarmour@redhat.com>

        * cdl/net.cdl (CYGHWR_NET_DRIVER_ETH0): Make flavor bool so it isn't
        defined when 0.
        (CYGHWR_NET_DRIVER_ETH1): Ditto.

2001-10-04  Hugo Tyson  <hmt@redhat.com>

        * src/lib/dhcp_support.c: Some conditional compilation tags were
        wrong vis a vis ETH1 versus ETH0.  Thanks to "Simon"
        <simoncc@ms46.url.com.tw> for spotting it.

2001-10-04  Jesper Skov  <jskov@redhat.com>

        * tests/multi_lo_select.c (cyg_user_start): Added CYG_TEST_INIT
        call.
        * tests/ping_lo_test.c (cyg_start): Same.
        * tests/tcp_lo_test.c (cyg_start): Same.
        * tests/udp_lo_test.c (cyg_start): Same.
        * tests/tcp_lo_select.c (cyg_start): Same.

2001-09-28  Jonathan Larmour  <jlarmour@redhat.com>

        * cdl/net.cdl: Tell isoinfra we support getproto* and getserv*
        NS functionality.
        * include/netdb.h: Move to...
        * include/net/netdb.h: ...here.

2001-09-25  Jesper Skov  <jskov@redhat.com>

        * cdl/net.cdl: Don't build tests/nc_test_master as it's a host
        side tool.

        * src/lib/getserv.c: Added domain service.

        * include/netdb.h: Moved getbyhost()/getbyaddr() declarations and
        hostent struct definition to the DNS package. 

        * src/lib/gethost.c: Removed.

        * cdl/net.cdl: Removed gethost.c (replaced by proper
        implementation in CYGPKG_NS_DNS).

2001-09-17  Hugo Tyson  <hmt@redhat.com>
2001-09-17  Andrew Lunn <Andrew.Lunn@ascom.ch>

        * src/sys/net/if_bridge.c (bridge_broadcast): Fix a null pointer
        deference.  Supporting VLANs, it seems that during the net_init()
        function, it sends a packet to the bridge with the source
        interface being NULL.  Stats counting assumed otherwise.

2001-09-14  Jonathan Larmour  <jlarmour@redhat.com>

        * src/sys/kern/uipc_syscalls.c: Last change to uipc_syscalls.c
        wasn't right. The parts that were __ECOS__ should simply have
        been completely removed.

2001-09-13  Hugo Tyson  <hmt@redhat.com>

        * cdl/net.cdl: Add configury to control whether we can tickle the
        network devices if there's a lack of network traffic, and to
        control the delay time before so doing.

        * src/ecos/timeout.c (alarm_thread): Use a timed wait, if so
        configured, for the flag; and if it times out, tickle all the
        devices via the common ether driver routine for so doing.

2001-09-12  Jonathan Larmour  <jlarmour@redhat.com>

        * include/sys/socketvar.h: Use __ECOS, not __ECOS__.
        * src/sys/kern/uipc_syscalls.c: Ditto.
        * src/sys/kern/uipc_socket2.c (sblock): Ditto.

2001-08-02  Jonathan Larmour  <jlarmour@redhat.com>

        * tests/linux_echo.c (echo_test): Set socket options before bind.
        * tests/nc_test_slave.c (do_tcp_test): Ditto.
        * tests/server_test.c (server_test): Ditto.
        * tests/tcp_echo.c (echo_test): Ditto.
        * tests/tcp_lo_test.c (server): Ditto.
        * tests/tcp_sink.c (sink_test): Ditto.
        * tests/tcp_source.c (source_test): Ditto.

2001-07-27  Jonathan Larmour  <jlarmour@redhat.com>

        * include/sys/select.h: Renamed to....
        * include/sys/bsdselect.h: New file.
        * include/sys/socketvar.h: Include <sys/bsdselect.h> instead of
        <sys/select.h>
        * src/lib/select.c: Include <sys/bsdselect.h>

2001-07-26  Jonathan Larmour  <jlarmour@redhat.com>

        * cdl/net.cdl (CYGPKG_NET_API_LOCAL): Implements select()
        * include/sys/bsdtypes.h: No need for these select definitions.

2001-06-20  Grant Edwards <grante@visi.com>
2001-06-20  Hugo Tyson  <hmt@redhat.com>

        * src/lib/dhcp_prot.c (alarm_function): Change the lease state
        before re-enabling the alarm so that if it somehow gets times of
        zero (ie. right now) it quickly completes the state machine rather
        than recursing to its doom.
        (new_lease): Test the retcode of get_bootp_option() and use
        obvious defaults if the T1 and T2 times are not provided - this is
        RFC compliant! - and use "infinite lease" if the lease time is not
        provided at all.

2001-06-19  Trenton D. Adams  <tadams@extremeeng.com>

        * tests/server_test.c (server_test): Null terminate read string
        at right place.

2001-06-13  Jonathan Larmour  <jlarmour@redhat.com>

        * cdl/net.cdl: Make debug output an option. Include DHCP.

        * src/lib/dhcp_prot.c: DHCP_CHATTER -> CYGDBG_NET_DHCP_CHATTER

        * src/ecos/support.c (cyg_kmem_init): Only print debug output if
        requested.

        * cdl/net.cdl: Make requirements on other eCos features more
        abstract.

2001-05-09  Robin Farine <acnrf@dial.eunet.ch>
2001-05-09  Hugo Tyson  <hmt@redhat.com>

        * src/lib/dhcp_prot.c (do_dhcp): In DHCPSTATE_INIT case, create a
        new xid.  Servers apparantly can use *only* this to distinguish
        machines, even with different MAC addresses!  Therefore we use
        both the most sensitive randomizer available (arc4random()) which
        in reality uses a finegrain clock, and salt the value further with
        the MAC address itself.  Thanks again Robin.

        * src/ecos/support.c (arc4random): Make arc4random not always be a
        multiple of 256; stir the clock into the low bits also.

2001-05-09  Robin Farine <acnrf@dial.eunet.ch>
2001-05-09  Hugo Tyson  <hmt@redhat.com>

        * src/sys/net/if_bridge.c: Patch from Robin; the route aging
        process took twice as long to timeout because an entry requires
        *two* executions of bridge_rtage() to actually get removed.
        Record the timeout over 2 to fix that and also define
        BRIDGE_RTABLE_TIMEOUT as 300s that the standard specifies as
        default aging time.

2001-05-09  Hugo Tyson  <hmt@redhat.com>

        * src/lib/getserv.c (setreturned): New routine to set up a copy of
        the service data with the port converted to network order like the
        API demands.  It's this way to make the initialization of the
        table most readable.  
        (services[]): Added an entry for snmp.
        (getservbyname): Use setreturned() to return adjusted info.
        (getservbyport): Use setreturned() to return adjusted info and
        compare with host-ordered version of the port.

        * src/lib/tftp_client.c (tftp_get): 
        (tftp_put): Do not convert to net order from
        the getserv structure; it's already net order.

        * src/lib/tftp_server.c (tftpd_server): Convert the socket into
        host order; it's network order in the getserv API.

        * tests/ftp_test.c (ftp_test): Do not convert to net order from
        the getserv structure; it's already net order.

        * tests/tftp_client_test.c (PUTFILE): Changed the names of the
        test files to be fully qualified, this makes the test work with
        more server machines.  /tftpboot/tftp_get and /tftpboot/tftp_put.
        
2001-04-24  Bart Veer  <bartv@redhat.com>

        * cdl/net.cdl:
        Prevent multiple device drivers from implementing the
        same eth0/eth1 device.

2001-03-30  Jonathan Larmour  <jlarmour@redhat.com>

        * cdl/net.cdl (CYGPKG_NET_SYSCTL): Comment out - currently unsupported.

2001-03-28  Richard Panton  <rpanton@3glab.com>

        * include/bootp.h: Define func protos with C linkage
        * include/netdb.h: Ditto
        * include/network.h: Ditto
        * include/tftp_support.h: Ditto

        * src/lib/getserv.c: getservbynumber() -> getservbyport()       

2001-03-12  Gary Thomas  <gthomas@redhat.com>

        * src/lib/network_support.c (init_all_network_interfaces): 
        FIx slightly broken previous attempt in abort code.

2001-03-11  Gary Thomas  <gthomas@redhat.com>

        * src/lib/network_support.c (init_all_network_interfaces): 
        Let initialization of PCMCIA devices abort after ~5 seconds.

2001-02-23  Jonathan Larmour  <jlarmour@redhat.com>

        * src/lib/tftp_server.c (STACK_SIZE): Align

2001-02-20  Jonathan Larmour  <jlarmour@redhat.com>

        * cdl/net.cdl: Avoid puts to the CDL header when a CDL
        define will do.
        Rename CYGPKG_NET_BRIDGE_CODE -> CYGPKG_NET_BRIDGE
        Rename CYGPKG_NET_NBRIDGE -> CYGNUM_NET_BRIDGES
        Rename CYGPKG_NET_BRIDGE_HANDLER -> CYGINT_NET_BRIDGE_HANDLER
        Make if_bridge.c compilation conditional on CYGPKG_NET_BRIDGE

        * src/sys/net/if_bridge.c: Reflect above renames.

2001-02-20  Hugo Tyson  <hmt@redhat.com>
2001-02-15  Martin Buck  <martin.buck@ascom.ch>

        * cdl/net.cdl: Split up CYGPKG_NET_NBRIDGE into 2 parts:
        CYGPKG_NET_NBRIDGE and CYGPKG_NET_BRIDGE_HANDLER. If
        CYGPKG_NET_BRIDGE_HANDLER is nonzero, calls to an Ethernet
        bridge are inserted into the stack's data path. If
        CYGPKG_NET_NBRIDGE is nonzero, the standard OpenBSD bridge
        code is added. This setup allows you to implement your own
        bridge as a separate package which then implements
        CYGPKG_NET_BRIDGE_HANDLER so it gets called by the stack.
        NBRIDGE now is the same as CYGPKG_NET_BRIDGE_HANDLER instead
        of CYGPKG_NET_NBRIDGE. This means it's nonzero if the bridge
        is wanted, but it no longer contains the number of bridge
        buffers requested. This shouldn't be a problem, because the
        only place where the actual number of buffers is required is
        in if_bridge.c, which now uses CYGPKG_NET_NBRIDGE instead
        of NBRIDGE.

        * src/sys/net/if_bridge.c: Use CYGPKG_NET_NBRIDGE instead of
        NBRIDGE, because the latter one now only means that there is
        a bridge, not that we should implement it.

2001-01-07  Gary Thomas  <gthomas@redhat.com>

        * src/ecos/support.c: 
        * cdl/net.cdl:  Add interface 'CYGPKG_NET_DRIVER_FRAMEWORK'
        to describe interdependencies between network stack and driver
        framework packages.


2001-01-03  Hugo Tyson  <hmt@redhat.com>

        * tests/linux_echo.c: New file to test for network bandwidth
        limitations in host side.  Builds to a linux version of the
        tcp_echo middleman.

        * tests/make.linux (all): Build linux_echo.

2001-01-03  Hugo Tyson  <hmt@redhat.com>

        * include/dhcp.h (struct dhcp_lease): Define the semaphore pointer
        field in each lease structure.

        * src/lib/dhcp_support.c: Initialize the dhcp_lease structs to
        point to the semaphore "dhcp_needs_attention".

        * src/lib/dhcp_prot.c (alarm_function): Post to the semaphore
        pointed to in the lease structure, rather than one hard-coded.
        This is much cleaner, in that dhcp_prot.c now uses no external
        variables, all state goes through the API.

2000-11-15  Hugo Tyson  <hmt@redhat.com>

        * src/lib/bootp_support.c (init_net): Do a SIOCSIFADDR a 2nd time
        after setting the netmask (SIOCSIFNETMASK) in order for the newly
        set netmask to "take" - otherwise a bogus route based on the
        default netmask lurks within the system.

2000-11-10  Hugo Tyson  <hmt@redhat.com>

        * src/ecos/support.c (cyg_net_get_mem_stats): New API for getting
        info on the various mem pools the stack uses to enable tests to
        spot store leaks.

        * include/network.h (cyg_net_get_mem_stats): Export this API for
        automated network testing.

2000-10-24  Hugo Tyson  <hmt@redhat.com>

        * src/lib/tftp_server.c (tftpd_server): Cut down the chatter to
        nothing (unless there's an error) if running automated testing.

2000-10-17  Hugo Tyson  <hmt@redhat.com>
2000-10-10  Andrew Lunn <Andrew.Lunn@ascom.ch>

        * src/ecos/support.c (setsoftnet,cyg_panic): Less chatter,
        particularly not when we're out of MBUFs.

2000-10-17  Hugo Tyson  <hmt@redhat.com>
2000-10-10  Andrew Lunn <Andrew.Lunn@ascom.ch>

        * src/sys/net/if_bridge.c: Intergrated a more up to date version
        from the OpenBSD sources. This fixes a few bugs.

2000-10-17  Hugo Tyson  <hmt@redhat.com>

        * src/ecos/support.c (cyg_ktime_init): Start time at 1 Second so
        that all visible time values are valid; offset time by 1 Second
        subsequently, for ever.

        * src/sys/netinet/if_ether.c (arpresolve): Undo the change below;
        instead make "kernel time" be valid ie. more than 1 second into
        its life.  Dumb stack assumes it takes UNIX-like time to start.

2000-10-16  Hugo Tyson  <hmt@redhat.com>

        * src/sys/netinet/if_ether.c (arpresolve): Deal with the initial
        case when the route timeout is zero (as initialized).  Otherwise
        it never actually sent out the initial ARP request packet.  This
        caused connect() not to work, and thus ftp_test, for example.

        * src/lib/dhcp_support.c (dhcp_start_dhcp_mgt_thread): Also
        initialize the dhcp semaphore here.  The other place it is init'd
        [correctly] didn't fire if we use the standard support.  Fixed
        some absent return value warnings.

2000-10-10  Hugo Tyson  <hmt@redhat.com>

        * src/lib/tftp_server.c (tftpd_server): Modify the server to
        support multiple sessions - ie. starting N servers at once.
        Mainly this means closing the initial socket whilst servicing a
        request, so that another server can then bind to it; tell another
        server to retry that bind via a semaphore, that it waited on when
        the bind failed initially, rather than just returning.

        * src/lib/tftp_dummy_file.c (dummy_open): Trivial bugfix: scan the
        list of files as well as incrementing the counter.

2000-10-05  Andrew Lunn <andrew.lunn@ascom.ch>

        * src/ecos/support.c (cyg_ktime_func,cyg_ktime_init,cyg_net_init): 
        Make 'ktime' value valid.

2000-10-05  Hugo Tyson  <hmt@redhat.com>

        * src/lib/tftp_server.c (tftpd_write_file): Restructure this
        function to match the loop layout of the tftpd_read_file; so that
        it retries sensibly, and so that a delayed/duplicated packet does
        not cause a doubling of all traffic with a less smart host.

2000-10-05  Hugo Tyson  <hmt@redhat.com>

        * include/tftp_support.h (TFTP_TIMEOUT_MAX): Change this to 50; it
        refers to total timeouts for a whole session, so it should be
        greater than the retry count for each packet.

2000-10-05  Hugo Tyson  <hmt@redhat.com>

        * src/lib/tftp_dummy_file.c: Make the fake file slots be a Mb
        instead of 10k so that you can do meaningful timing tests with it.
        (dummy_open): Allow re-write of an existing file so that you can
        do repeated put tests without running out of slots.

2000-10-05  Hugo Tyson  <hmt@redhat.com>

        * src/lib/select.c (_cyg_select): Unlock the scheduler in a couple
        of other places I missed.  Doh.

2000-09-28  Hugo Tyson  <hmt@redhat.com>

        * src/lib/select.c (_cyg_select): Elect to wait for the flags
        atomically wrt sockets possibly becoming ready - this was a race
        condition that would apparently delay a packet until another
        arrived, eg. a tftp retry.  Then two came along all at once.

2000-09-28  Hugo Tyson  <hmt@redhat.com>

        * src/lib/tftp_server.c: Add lots of instrumentation for
        debugging.  It's switched off, and doesn't have a real config opt,
        though it easily could have.

2000-09-26  Hugo Tyson  <hmt@redhat.com>

        * src/lib/tftp_server.c (tftpd_read_file): Doh! TFTP_TIMEOUT is an
        internal API thing, not an error we can send in an ERROR packet.

        * include/tftp_support.h: Comment to this effect.

2000-09-26  Hugo Tyson  <hmt@redhat.com>
2000-09-25  Andrew Lunn <Andrew.Lunn@ascom.ch>

        * src/lib/tftp_server.c: Send an ERROR packet when giving up after
        too many timeouts.  This should cause the client to give up as
        well.  Also moved all the replicated code to send an ERROR packet
        into one function.

        [Huge] I collected another point where we can use the common
        function also, and used ETIMEOUT instead of EBADOP for the new
        error packet returns. Thanks Andrew!
        
2000-09-14  Hugo Tyson  <hmt@redhat.com>

        * cdl/net.cdl (CYGOPT_NET_DHCP_DHCP_THREAD_PARAM): Set default to
        1 so that the DHCP management thread loops forever.  This allows
        it to recover from a DHCP renewal failure.

        * src/lib/dhcp_prot.c (next_timeout): Harden the DHCP protocol
        machine against [simulated] failures - otherwise the tests with
        simulated failures tend just to close down all the interfaces when
        DHCP packets are lost.

2000-09-11  Gary Thomas  <gthomas@redhat.com>

        * src/lib/bootp_support.c (init_net): Set default route correctly.

2000-09-01  Hugo Tyson  <hmt@cygnus.co.uk>

        * src/ecos/support.c (cyg_net_init): You can't print things while
        initializing the network!  Well, not if connected to GDB over the
        network anyway.  The printf("Init device '%s'...); removed.

2000-09-01  Hugo Tyson  <hmt@cygnus.co.uk>

        * OVERVIEW: This is part of the change to the network stack to
        greatly reduce latencies both of (other) DSRs and of thread
        scheduling.  All the work that the network stack *and* individual
        ether drivers used to do in DSRs (including alarm callbacks and
        data copies to/from the device memory) is moved into a "fast
        network thread" instead.  It calls a device's "deliver" function
        to do the work that was previously in the DSR.  This is a separate
        thread so that it can be set higher priority than application
        threads in order to minimize packet loss (depending on the
        driver), if required (the application threads presumed to be
        higher priority in turn than the network thread).  A crucial
        consequence of this is that we are no longer locking against DSRs,
        so a plain mutex can be used rather than the global scheduler
        lock, thus simplifying all the splfoo/splx() style functions.

        * src/ecos/timeout.c (alarm_thread): Addition of the "fast network
        thread" which runs DSR-like activities.
        (do_timeout): Timeout function morphed for calling from that.
        (do_alarm, ecos_synch_eth_drv_dsr): new DSR functions to signal to
        the thread.
        (timeout): Race condition fixed.  splinternal() used for locking
        instead of scheduler.
        
        * src/ecos/support.c (cyg_net_init): Splfoo/splx() functions,
        together with tsleep/wakeup functions, all removed to separate
        them from the mixed bag of utilities in this file.  What remains
        is mbuf wrapper routines and the like, plus the network "netisr"
        thread itself, the caller into the stack that does the slower
        priority work.

        * src/ecos/synch.c: New file; implemtation of new splfoo/splx()
        functions, together with tsleep/wakeup functions, since they are
        related now.

        * cdl/net.cdl: Compile new file synch.c; two new options, one for
        "fast thread" priority, and one for DHCP manager thread priority,
        as I was adding prio configury.  CYGPKG_NET_FAST_THREAD_PRIORITY
        and CYGPKG_NET_DHCP_THREAD_PRIORITY resp, with suitable default
        values relative to the CYGPKG_NET_THREAD_PRIORITY.

        * src/lib/dhcp_support.c (dhcp_start_dhcp_mgt_thread): Use the
        configured priority rather than just "net thread - 1"

2000-08-31  Hugo Tyson  <hmt@cygnus.co.uk>

        * tests/tcp_echo.c: Change the priorities of the main and loading
        threads to accommodate the network having helper threads around at
        adjacent priorities to its main thread prio.

2000-08-24  Hugo Tyson  <hmt@cygnus.co.uk>

        * src/sys/net/if.c (ifioctl): Support the two new ioctl() keys; at
        this level, the struct ifreq at the head of the data area must be
        filled to select an interface.

        * include/sys/sockio.h (SIOCGIFSTATSUD): Add two new eCos-only
        ioctl() keys, SIOCGIFSTATSUD and SIOCGIFSTATS, for reading
        statistical information out of ethernet devices, for SNMP. This
        should allow SNMP (a) to not explode, (b) to get useful info out
        of other device implementations than this one.

2000-08-17  Hugo Tyson  <hmt@cygnus.co.uk>

        * src/ecos/timeout.c (timeout): Rework the timeout system to
        record last-time-we-set-the-alarm and whence values, so that new
        timeouts being added can be set up relative to the correct
        absolute time.  Otherwise adding a new timeout sets them *all*
        into the future by the expired portion of the previous minimum
        pending timeout.  Also deal better with recursion ie. timeout
        handlers themselves setting new timeouts as is only natural.
        Lots of asserts too.

        * src/ecos/support.c (cyg_splsoftnet): Use the new kernel facility
        to lock mutex &c atomically, with the scheduler locked already.
        (cyg_tsleep): Similarly, and reclaim the mutex likewise.
        Also added lots of asserts to both calls the better to document
        what's going on.

2000-08-17  Hugo Tyson  <hmt@cygnus.co.uk>

        * src/lib/tftp_server.c: Fix contributors field.

2000-08-16  Hugo Tyson  <hmt@cygnus.co.uk>

        * src/lib/tftp_server.c: Improvements to server wrt better error
        messages and detection of filesystem errors eg. on close-file.
        Contrib from ASCOM.  I tidied up some comments and indent to
        minimize the diff.  [CASE 104354]

2000-08-15  Hugo Tyson  <hmt@cygnus.co.uk>

        * doc/bridge.html: New file...
        * doc/bridge.doc: New file...
        provided by ASCOM from the OpenBSD version.

2000-08-15  Hugo Tyson  <hmt@cygnus.co.uk>

        * src/sys/net/if_bridge.c (bridge_ioctl): SIOCBRDGSTO takes
        argument in seconds not in ticks.  So move around some mul/div by
        hz ops to get this right.

2000-08-14  Gary Thomas  <gthomas@redhat.com>

        * src/ecos/support.c: Use new table definition mechanism.

2000-07-27  Hugo Tyson  <hmt@cygnus.co.uk>

        * src/lib/dhcp_prot.c (do_dhcp): Use xmit->bp_htype =
        HTYPE_ETHERNET rather than ifr.ifr_hwaddr.sa_family; sa_family is
        in a different namespace, despite appearances.

        * src/lib/bootp_support.c (do_bootp): Use bp_htype =
        HTYPE_ETHERNET rather than ifr.ifr_hwaddr.sa_family; sa_family is
        in a different namespace, despite appearances.

2000-07-27  Nick Garnett  <nickg@cygnus.co.uk>

        * cdl/net.cdl: Require the C library STDIO package to be present
        if there is more than one loopback interface. This is because it
        needs sprintf() to form the interface names.

        * src/sys/net/if_loop.c (loopattach): 
        Only use sprintf() to form the loopback interface names when there
        is more than one. The CDL ensures that this will work.

        * src/lib/network_support.c:
        Added diag_printf() version of perror() for when STDIO package is
        absent.
        Only use sprintf() to form the loopback interface names when there
        is more than one. The CDL ensures that this will work.

        * tests/multi_lo_select.c:
        * tests/ping_lo_test.c:
        * tests/tcp_lo_test.c:
        Removed dependence on C library by adding a substitute perror()
        based on diag_printf() that is enabled when the STDIO package is
        absent.
        
2000-07-26  Hugo Tyson  <hmt@cygnus.co.uk>
 
        * src/lib/dhcp_prot.c (do_dhcp): Set the broadcast flag where
        necessary, and the client address too.  This makes it talk to a
        greater range of servers OK.
 
2000-07-25  Nick Garnett  <nickg@cygnus.co.uk>

        * tests/multi_lo_select.c: Changed definition of NLISTENERS to
        work correctly with FILEIO package.

        * src/sys/net/if_loop.c (loopattach): Changed initialization of
        if_xname to "lo0".

        * src/ecos/support.c (cyg_net_init): Removed redundant code to
        initialize loop-back interface.

2000-07-25  Hugo Tyson  <hmt@cygnus.co.uk>

        * src/lib/getserv.c: Add a list terminator record with NULLs in it
        so that the search finishes (without throwing asserts about bad
        strings).

2000-07-21  Hugo Tyson  <hmt@cygnus.co.uk>

        * src/sys/net/if_bridge.c: Move the include of stdio.h lower,
        apparently it confuses local x86 compilers if their host tree is
        malformed.  Or something.

2000-07-21  Hugo Tyson  <hmt@cygnus.co.uk>

        * cdl/net.cdl: Add a lot of description about the way the various
        fields are used in interface initialization: specifically that
        "server" so-called is just "someone to talk to" in the absence of
        bootp - and our tests depend on it!

        * src/lib/bootp_support.c (init_net): Only set up a route if the
        route address is nonzero - or all manner of confusion occurs with
        multiple interfaces and route that sends to 0.0.0.0...
        For setting up the default route, use
              (SIOCADDRT, 0, 0, TAG_GATEWAY)
        rather than the bogus
              (SIOCADDRT, yiaddr & netmask, netmask, TAG_GATEWAY)
        that we did before.

        * tests/ftp_test.c (ftp_test): Try it with eth1 if available also.

        * tests/ping_test.c (ping_host): If a ping fails, reset the packet
        size to small just in case the huge packet size is what's causing
        confusion - helps it as a debug tool.

        * tests/dhcp_test.c (net_test): It didn't compile if DHCP is
        disabled!  Doh.  Same change as ping_test also.

2000-07-19  Hugo Tyson  <hmt@cygnus.co.uk>

        * src/lib/tftp_server.c (tftpd_read_file): Deal with a) [assumed]
        timeouts on the select, and b) ACKs for old packets.
        [CASE 104052 and CASE 104055]

2000-07-19  Hugo Tyson  <hmt@cygnus.co.uk>

        * src/lib/select.c (_cyg_select): Return 0 and do not perturb
        errno when the timeout occurs.  That matches the man page!
        [CASE 104054]

2000-07-19  Hugo Tyson  <hmt@cygnus.co.uk>

        * include/machine/param.h: Be more defensive against warnings from
        external defines such as __linux__/__bsdi__/__FreeBSD__/...
        [CASE 104090]

2000-07-19  Hugo Tyson  <hmt@cygnus.co.uk>

        * tests/multi_lo_select.c: New test program to test for proper
        broadcast behaviour of select() implementation.  Took some
        fiddling to get it to work, but it really did fail before the
        select change below.

        * cdl/net.cdl: Build the new test.

        * src/lib/select.c (_cyg_select): Don't use the CLR flag in the
        flag wait because that gives unicast semantics.  Flags have
        producer-does-all-the-work behaviour, so setting a value then
        clearing it right afterwards does the right thing, with broadcast
        semantics so long as no waiter has set the CLR part.
        [CASE 104058]

2000-07-18  Hugo Tyson  <hmt@cygnus.co.uk>

        All part of the bridge contribution from Andrew Lunn/ASCOM
        (andrew.lunn@ascom.ch).  Bridginess is controlled by NBRIDGE which
        is itself controlled by CDL option CYGPKG_NET_NBRIDGE aka "Number
        of bridge buffers?"

        * include/machine/param.h: Add proper definition of untimeout().

        * include/sys/sockio.h (SIOCBRDGFRL): Add this and the other 2
        bridge ioctl() call definitions. 

        * include/sys/param.h (splhigh): Added.

        * src/ecos/timeout.c (untimeout): Implement this properly, it was
        never used before.

        * src/ecos/support.c: implement cyg_splhigh(), call bridgeintr()
        when it is scheduled and bridgeattach() if configured.

        * tests/bridge.c: New "test" file - implements an ethernet bridge.
        Contributed by Andrew Lunn/ASCOM, from the usual OpenBSD external
        source original.

        * include/net/if_bridge.h: New file, contributed by Andrew Lunn/
        ASCOM, from OpenBSD original.  (Actually did exist in eCos source
        tree but was not released)

        * src/sys/net/if_bridge.c: New file, contributed by Andrew Lunn/
        ASCOM, from OpenBSD original.

        * cdl/net.cdl: Build the new files.

2000-07-18  Gary Thomas  <gthomas@redhat.com>

        * src/lib/tftp_client.c (tftp_get): Fix edge condition when 
        penultimate block is full and end of file which implies that
        the last block has zero data bytes.

2000-07-18  Hugo Tyson  <hmt@cygnus.co.uk>

        * src/lib/dhcp_prot.c: Ignore NAK messages from the wrong server -
        we get these because the REQUEST for our chosen IP is broadcast,
        so other servers think we've asked to use the wrong IP.

2000-07-14  Hugo Tyson  <hmt@cygnus.co.uk>

        * include/bootp.h: export init_loopback_interface();

2000-07-14  Hugo Tyson  <hmt@cygnus.co.uk>

        * cdl/net.cdl: Add option controlling whether to have a separate
        DHCP lease management thread, and what its parameter is to be -
        which controls whether it loops or quits if a lease fails.

        * src/lib/dhcp_support.c (dhcp_start_dhcp_mgt_thread): Provide the
        DHCP management thread function and code to instantiate and start
        the thread if so configured.

        * src/lib/network_support.c (init_all_network_interfaces): Call
        the function that starts the DHCP management thread function.

        * include/dhcp.h: Declare the DHCP management thread function &c
        if so configured.

        * tests/dhcp_test.c (net_test): Only poll for need to rebind DHCP
        leases if there is no service thread to do the same.

2000-07-14  Hugo Tyson  <hmt@cygnus.co.uk>

        * include/bootp.h (build_bootp_record): Export this, no harm in
        it, and some folk want to use this rather than configured
        initialization or bootp/dhcp.
        Also commented the other APIs somewhat better.

        * src/lib/network_support.c (build_bootp_record): Unconditionally
        provide this; selective linking will look after it if unused.

2000-07-14  Hugo Tyson  <hmt@cygnus.co.uk>

        * src/lib/network_support.c (build_bootp_record): [Static
        configuration of interface parameters] Add the gateway into the
        options section with TAG_GATEWAY so that init_net() will pick it
        up and set up a route accordingly.

2000-07-13  Hugo Tyson  <hmt@cygnus.co.uk>

        * tests/dhcp_test.c:
        * tests/flood.c:
        * tests/ftp_test.c:
        * tests/nc_test_master.c:
        * tests/nc_test_slave.c:
        * tests/ping_lo_test.c:
        * tests/ping_test.c:
        * tests/server_test.c:
        * tests/set_mac_address.c:
        * tests/tcp_echo.c:
        * tests/tftp_client_test.c:
        * tests/tftp_server_test.c:
        Up the stack size to cope with full DHCP initialization in
        init_all_network_interfaces().
        
2000-07-13  Hugo Tyson  <hmt@cygnus.co.uk>

        * cdl/net.cdl: Add DHCP enable and interface-specific controls.

        * include/dhcp.h: New file: describe the DHCP APIs and behaviour,
        as well as some internal APIs that are also available.

        * src/lib/dhcp_support.c: New file: the surrounding management
        routines, a bit like network_support.c, which know about multiple
        interfaces and so on.

        * src/lib/dhcp_prot.c: New file: the interface-independent DHCP
        protocol machine, which is called from dhcp_support.c routines and
        init_all_network_interfaces() in network_support.c

        * include/bootp.h (struct bootp): Make the packet bigger if DHCP
        so that a mininal DHCP packet will fit.

        * src/lib/bootp_support.c (do_bootp): If no reply, fail gracefully
        and tidy up so that other interfaces can be upbrung.
        (show_bootp): Tell us a lot more about DHCP-specific info in the
        packet structure.
        (get_bootp_option): be aware of DHCP extensions to re-use file and
        sname fields for options.

        * src/lib/network_support.c (init_all_network_interfaces): Call
        do_dhcp() rather than do_bootp() if so configured, and initialize
        the per-interface DHCP state variables accordingly.

        * tests/dhcp_test.c (net_test): New test file that diddles the
        DHCP machine while doing a ping test.

2000-07-11  Gary Thomas  <gthomas@redhat.com>

        * include/netdev.h: Add single-inclusion fences.

        * include/sys/param.h: Add traced versions of splx() routines,
        selectable by a configuration option.

        * src/ecos/support.c: Rework splx() emulation routines to
        be more robust and realtime friendly.

        * src/lib/network_support.c (init_all_network_interfaces): 
        Add minimal support for PCMCIA based devices.

2000-07-04  Hugo Tyson  <hmt@cygnus.co.uk>

        * include/bootp.h: Fix namespace pollution from "#define int32
        int" - it fights against application code too much.

2000-07-04  Jonathan Larmour  <jlarmour@redhat.co.uk>

        * cdl/net.cdl: Package requires CYGPKG_MEMALLOC

2000-06-26  Hugo Tyson  <hmt@cygnus.co.uk>

        * tests/ping_test.c (net_test): Added use of the
        CYGPKG_NET_TESTS_USE_RT_TEST_HARNESS stuff now that the test
        passes (consequent on the change below); also use larger ping
        packets now that that is working also.

        * tests/tcp_echo.c (echo_test): Added use of the
        CYGPKG_NET_TESTS_USE_RT_TEST_HARNESS stuff now that the test
        passes (consequent on the change below).

2000-06-26  Hugo Tyson  <hmt@cygnus.co.uk>

        * src/ecos/support.c (cyg_splnet): Use the scheduler lock and a
        mutex instead of disable-interrupts for SPLX type processing.  A
        mutex is used at splsoftnet because that is client threads - we do
        not want them to pre-empt the rest of the app.  This enables the
        real-time response testing for the EBSA285 to succeed (interrupts
        every 1mS, DSRs delayed by at most 2mS).

2000-06-23  Hugo Tyson  <hmt@cygnus.co.uk>

        * src/ecos/support.c (cyg_net_mbuf_alloc, cyg_kmem_init): Align
        the mbuf pool to MSIZE [128] bytes.  That way dtom() works, nasty
        though it is.  That's needed for ip reassembly in ip_input.c, when
        dealing with large icmp-layer packets eg. ping -s 2000 ...

2000-06-21  Hugo Tyson  <hmt@cygnus.co.uk>

        * include/lib/libkern/libkern.h: Do not define assert multiple
        times; guarded by __ECOS.

        * include/sys/param.h (MAX,MIN): Only define if undef.

2000-06-21  Hugo Tyson  <hmt@cygnus.co.uk>

        * cdl/net.cdl: Build the new tests; in fact build them
        unconditionally.

        * src/lib/network_support.c (init_all_network_interfaces):
        Initialize loopback device[s] using init_loopback_interface(),
        providing a sensible default route et al.  Multiple devs: lo1 will
        be 127.0.*1*.1, and so on; class C netmasks will be used.

        init_all_network_interfaces() is also made thread-safe as well as
        idempotent, so other threads (or daemons) can call it to make sure
        the net is up.

        * src/ecos/support.c (cyg_net_init): Add a call to loopattach() if
        there are indeed loopback dev(s) configured.  This does the
        equivalent of the init of a device from the table, but simpler.

        * tests/udp_lo_test.c (udp_server): New testcase...
        * tests/tcp_lo_select.c (tcp_server): New testcase...
        * tests/tcp_lo_test.c (tcp_client): New testcase...
        * tests/ping_lo_test.c (net_test): New testcase, unconditionally
        built loopback device test.  Will run on platforms with no network
        interfaces.

2000-06-19  Nick Garnett  <nickg@cygnus.co.uk>

        * src/ecos/support.c:
        * include/netdev.h:
        Converted to use of new table construction mechanism.

2000-06-16  Jonathan Larmour  <jlarmour@redhat.co.uk>

        * include/bootp.h: Remove RCS id tag
        * include/netinet/ip_auth.h: Ditto
        * include/netinet/ip_fil.h: Ditto
        * include/netinet/ip_fil_compat.h: Ditto
        * include/netinet/ip_frag.h: Ditto
        * include/netinet/ip_nat.h: Ditto
        * include/netinet/ip_proxy.h: Ditto
        * include/netinet/ip_state.h: Ditto

2000-06-15  Nick Garnett  <nickg@cygnus.co.uk>

        * src/sys/kern/sockio.c: Added cyg_selinit() calls to bad_socket()
        and bsd_accept() functions.

        * src/ecos/support.c (cyg_tsleep): Changed tests on wakeup from
        semaphore waits to enable return of EINTR results.

2000-06-09  Nick Garnett  <nickg@cygnus.co.uk>

        * src/sys/net/if_loop.c: Substituted sprintf() for a straight
        strcpy() when initalizing if name. This currently only works for
        one loopback interface. However, there seems little need to have
        more than one.

        * src/sys/kern/sockio.c (bsd_select): Added select support.

        * include/sys/time.h: Moved timeval structure definition to
        isoinfra time.h header since it is needed by the select() API.

        * include/sys/select.h: Added option to use fileio select
        mechanism if it is present.

        * src/ecos/support.c (cyg_net_init):
        Added code to bring up the loopback interface. This is to help
        with testing on platforms without network hardware.

2000-06-08  Hugo Tyson  <hmt@cygnus.co.uk>

        * include/netinet/ip_var.h: Export ipforwarding as well as
        ip_defttl for monitoring applications.

2000-06-07  Nick Garnett  <nickg@cygnus.co.uk>

        * tests/tcp_echo.c: 
        * tests/socket_test.c: 
        * tests/server_test.c: 
        * tests/ping_test.c: 
        * tests/nc_test_master.c: 
        * tests/nc_test_slave.c: 
        * tests/ftp_test.c:
        Removed dependence of these programs on STDIO. For most this
        simply required the use of a diag_printf() based perror() clone is
        CYGPKG_LIBC_STDIO is not defined. For server_test is also required
        the substitution of sprintf() with some more primitive string
        functions.

        * src/sys/netinet/ip_input.c: Rewrote inet_ntoa() to not use
        sprintf(). This was the only place in the TCP/IP stack that was
        dependent on a STDIO function. We really should avoid building
        this kind of casual dependency into the code.

        * src/sys/kern/sockio.c: This new file contain code to support the
        fileio package. So far only a few of these functions have actually
        been tested.

        * include/sys/kernel.h (time): The time variable clashes with the
        C library time() function. To prevent this it is renamed ktime,
        and a #define allows "kernel" code to continue accessing it as
        time.

        * src/ecos/support.c: Added definition for ktime variable, as
        described above. I could not find any definition for struct
        timeval time. I have a nasty suspicion that it was using the
        time() function - let's hope that no code that actually uses this
        has been run.

        * src/lib/bootp_support.c: Another substitution of a diag_printf()
        based perror() clone when STDIO is absent.
        
        * include/sys/bsdtypes.h: This is the original
        include/sys/types.h. Renamed to avoid clashing with <sys/types.h>
        defined by the isoinfra package.

        * include/sys/types.h: Renamed to bsdtypes.h.
        
        * include/machine/limits.h:
        Various changes to make these headers play nicely with those
        defined in isoinfra package.

        * include/network.h: Added include of <pkgconf/system.h>. Fixed
        prototypes of functions here to match POSIX specifications.

        * cdl/net.cdl:
        Split list of files to compile into the common set plus those
        needed when the fileio package is present and those that are
        needed when it is absent.
        Added CDL to export our definitions to <sys/types.h>.
        

2000-05-31  Hugo Tyson  <hmt@cygnus.co.uk>

        * src/lib/network_support.c (init_all_network_interfaces): Make
        this call idempotent - this is useful for reliable initialization
        of dependent subsystems.

        * include/lib/libkern/libkern.h: remove protos of random() and
        srandom() for they are not in fact provided.

        * cdl/net.cdl: Turn off CYGPKG_NET_TESTS_USE_RT_TEST_HARNESS by
        default - since not all tests currently pass. 

2000-05-12  Hugo Tyson  <hmt@cygnus.co.uk>

        * cdl/net.cdl (CYGPKG_NET_TFTPD_THREAD_PRIORITY): New option,
        control the TFTPD thread priority.  These options are *all*
        CYGPKG_NET.  Yuk, sort 'em out later.

        * tests/tftp_server_test.c (tftp_test): Added use of the
        CYGPKG_NET_TESTS_USE_RT_TEST_HARNESS stuff, same as for the flood
        ping test.  This one passes (with the next change), so committed.

        * src/lib/tftp_server.c (tftpd_server): Cancelled some of the
        printouts as connections come and go *iff* the tests are set up to
        use the realtime-ness test harness.  This needs generalizing into
        proper control of the network's chattiness overall.
        (tftpd_start): Also added configury of the TFTPD thread priority.

2000-05-12  Hugo Tyson  <hmt@cygnus.co.uk>

        * cdl/net.cdl: Add flood test below.  Also add option
        CYGPKG_NET_TESTS_USE_RT_TEST_HARNESS to decide whether to
        "Use real-time response test harness (if available)".
        Consequently moved the test build stuff forward out of the build
        flags area.

        * tests/flood.c (net_test): New file; performs a flood-ping (well,
        as fast as we can go) of the server(s) on the two interfaces.
        Also uses the real-time interrupt response verification stuff from
        the EBSA285 driver component.

2000-05-11  Hugo Tyson  <hmt@cygnus.co.uk>

        * cdl/net.cdl: Add new utility (and harmless-by-default test)
        tests/set_mac_address.  You have to edit it to get anything
        damaging to happen.

        * tests/set_mac_address.c: New file added.  It uses SIOCSIFHWADDR
        to set the MAC address of any interfaces it has been told to.

2000-05-08  Gary Thomas  <gthomas@redhat.com>

        * include/machine/cpu.h: 
        * include/machine/cdefs.h: Add copyright boilerplate.

2000-05-05  Hugo Tyson  <hmt@cygnus.co.uk>

        * include/arpa/tftp.h: Aargh.  Turns out that removing the align
        driective and retaining only the packed directive makes the
        original version work.  So, reverted.  Apologies for the wasted
        time.

        * src/lib/tftp_server.c (tftpd_read_file): Coupla warnings
        reduced, and reverted to match original tftp.h

        * src/lib/tftp_dummy_file.c (dummy_open): Removed use of
        undeclared malloc().

2000-05-04  Gary Thomas  <gthomas@redhat.com>

        * src/lib/tftp_server.c: 
        * src/lib/tftp_client.c: Adjust for changes in header structure.

        * include/arpa/tftp.h: Continuing problems with alignment on ARM.  
        Recourse is very bastardized structure, but it _does_ work.

2000-05-04  Hugo Tyson  <hmt@cygnus.co.uk>

        * tests/tftp_client_test.c (tftp_test): Print out a coupla more
        things and test both interfaces if they exist - which does the
        same thing twice if the same server bootp'd both.  Changed the
        filenames to something more obvious.

        * cdl/net.cdl (CYGPKG_NET_TESTS): Build the tftp tests
        tests/tftp_client_test tests/tftp_server_test

        * src/ecos/support.c (cyg_kmem_print_stats): New function; prints
        info about memory usage for some tests to come.

2000-05-04  Gary Thomas  <gthomas@redhat.com>

        * include/arpa/tftp.h: Force packed alignment - required on some
        architectures.

2000-05-01  Gary Thomas  <gthomas@redhat.com>

        * src/lib/tftp_dummy_file.c: Adding very simple routines - just
        enough to test TFTP server.  This support will come from elsewhere
        in actual application environments (e.g. a real file system).

        * tests/tftp_server_test.c: 
        * src/lib/tftp_server.c: 
        * include/tftp_support.h: Flesh out TFTP server support.

2000-04-13  Hugo Tyson  <hmt@cygnus.co.uk>

        * tests/tcp_echo.c (calibrate_load): Import better background
        thread loading algorithm with bugfix; if the initial HIGH limit
        was not large enough, it never got loaded enough.

        * tests/nc_test_slave.c (calibrate_load): Import the fixed version
        from above back in here.

2000-04-12  Gary Thomas  <gthomas@redhat.com>

        * src/ecos/support.c (cyg_tsleep): Use 'cyg_scheduler_safe_lock()' so
        this function can be called with the scheduler locked.

        * src/sys/kern/uipc_socket2.c: 
        * include/sys/socketvar.h: Update sblock()/sbunlock() to be eCos safe.

2000-04-12  Hugo Tyson  <hmt@cygnus.co.uk>

        * tests/tcp_echo.c: Include <lib/libkern/libkern.h> so that it
        links - otherwise max() is not around.

2000-04-11  Gary Thomas  <gthomas@redhat.com>

        * include/lib/libkern/libkern.h: 
        * include/network.h (NO_LIBKERN_INLINE): Disable kernel inline
        functions.

2000-04-11  Gary Thomas  <gthomas@redhat.com>

        * src/lib/getserv.c: Add TFTP protocol.

        * cdl/net.cdl: Add TFTP library functions.
        
        * include/tftp_support.h: 
        * src/lib/tftp_dummy_file.c: 
        * src/lib/tftp_server.c: 
        * src/lib/tftp_client.c: New file(s).  Basic TFTP support functions.

2000-04-10  Hugo Tyson  <hmt@cygnus.co.uk>

        * tests/tcp_echo.c (echo_test): 
        * tests/tcp_source.c: 
        * tests/tcp_sink.c: 
        Merge in changes from Grant Edwards <grante@visi.com> presented on
        ecos-discuss "Mon, 10 Apr 2000 11:01:54 -0500" - fix some
        warnings, and use ntohl() on the control data that travels netly
        so that the host tools will work on otherendian machines.

2000-04-07  Gary Thomas  <gthomas@redhat.com>

        * tests/nc_test_slave.c: Update background thread loading
        calibration to use a binary search (better) algorithm.

2000-03-29  Hugo Tyson  <hmt@cygnus.co.uk>

        * tests/ftp_test.c (cyg_test_exit): 
        * tests/mbuf_test.c (cyg_start): 
        * tests/nc_test_master.c (cyg_test_exit): 
        * tests/nc_test_slave.c (show_net_times): 
        * tests/ping_test.c (cyg_test_exit): 
        * tests/server_test.c (cyg_test_exit): 
        * tests/socket_test.c (cyg_test_exit): 
        * tests/tcp_echo.c (cyg_test_exit): 
        Some eCos infrastructure changes caused, in some configurations,
        cyg_test_exit() from the infrastructure to be brought in (in
        tcdiag.o) even when the Test Case system was not explicitly being
        invoked.  That fought with the convenience copy of cyg_test_exit()
        in these tests.  The neatest fix is to use that provided centrally
        where relevent, so that is what this change does.

        * tests/tcp_echo.c: Also now use granularity of 5% by default,
        calibrate the load at 50%, and check and report the load that was
        actually achieved after the test.

2000-03-29  John Dallaway  <jld@cygnus.co.uk>

        * doc/ecos_tcpip.html:

        Remove error-prone cross-reference to another eCos package.

        * doc/index.html:

        Synchronize index page title with web site.

2000-03-28  Gary Thomas  <gthomas@redhat.com>

        * src/lib/bootp_support.c (init_net): Need to start interface,
        especially for manual/static IP configurations.

        * tests/nc_test_slave.c (calibrate_load): Better calibration for
        slower targets.

2000-03-18  Gary Thomas  <gthomas@redhat.com>

        * src/sys/kern/sys_socket.c: 
        * src/sys/kern/sys_generic.c: 
        * include/sys/sockio.h: Add FIONBIO, FIOASYNC, FIONREAD functions.

        * include/machine/types.h: Make definitions safe for use with libc.

2000-03-08  Gary Thomas  <gthomas@redhat.com>

        * src/ecos/support.c: Remove some debug messages.
        Update timed sleep functions (tsleep) to use scheduler lock
        instead of brute-force interrupt locks.

2000-03-08  Hugo Tyson  <hmt@cygnus.co.uk>

        * tests/tcp_echo.c (echo_test): Also enable, log and print out the
        idle-thread activity during the test - this reassures us that
        there is no [significant] spare time being wasted idling during
        the test.  This change has no effect on the results at all.

2000-03-07  Hugo Tyson  <hmt@cygnus.co.uk>

        * tests/tcp_echo.c (calibrate_load): Improve the accuracy of the
        calibration with some post-scaling.  Change load to an array
        accessor instead of register-based floating point - this is less
        load, so more loops are needed, so the initial load level is upped
        also.  More loops means better accuracy anyway, a good thing.

2000-03-06  Hugo Tyson  <hmt@cygnus.co.uk>

        * tests/tcp_source.c (show_results): Print results in Mbit/S also.
        (source_test): Warnings reduction; add some casts to sending and
        receiving the control packets.
        (NUM_BUF): becomes 1024; send more data to get thro'put result.

        * tests/tcp_sink.c (show_results): Print results in Mbit/S also.
        (sink_test): Warnings reduction; add some casts to sending and
        receiving the control packets.

        * tests/nc_test_master.c (show_results): Reinstate tot_bytes
        variable for host testing tool version.  Print results in Mbit/S
        also.

        * tests/nc_test_slave.c (net_test): Comment out starting multiple
        threads; they fight over a wide-open socket.  One thread is
        enough.

2000-03-06  Gary Thomas  <gthomas@redhat.com>

        * include/netdev.h: Update structure for improved API description.

2000-03-05  Gary Thomas  <gthomas@redhat.com>

        * include/network.h:
        * include/sys/syscallargs.h:
        * src/lib/accept.c:
        * src/lib/bootp_support.c:
        * src/lib/close.c:
        * src/lib/getpeername.c:
        * src/lib/getsockname.c:
        * src/lib/recvfrom.c:
        * tests/ftp_test.c:
        * tests/nc_test_framework.h:
        * tests/nc_test_master.c:
        * tests/nc_test_slave.c:
        * tests/ping_test.c:
        * tests/server_test.c:
        * tests/socket_test.c:
        * tests/tcp_echo.c:
        Cleanup to remove compiler warnings.    

2000-03-04  Gary Thomas  <gthomas@redhat.com>

        * include/machine/param.h:
        * include/sys/socketvar.h:
        * src/ecos/timeout.c:
        * src/sys/kern/kern_subr.c:
        * src/sys/kern/sys_generic.c:
        * src/sys/kern/sys_socket.c:
        * src/sys/kern/uipc_mbuf.c:
        * src/sys/kern/uipc_socket2.c:
        * src/sys/kern/uipc_syscalls.c:
        * src/sys/net/if_ethersubr.c:
        * src/sys/net/if_loop.c:
        * src/sys/net/route.c:
        * src/sys/net/rtsock.c:
        * src/sys/netinet/in_cksum.c:
        * src/sys/netinet/in_pcb.c:
        * src/sys/netinet/ip_input.c:
        * src/sys/netinet/tcp_input.c:
        * src/sys/netinet/tcp_output.c:
        * src/sys/netinet/tcp_subr.c:
        * src/sys/netinet/udp_usrreq.c:
        Cleanup to remove compiler warnings.    

2000-03-01  Gary Thomas  <gthomas@cygnus.co.uk>

        * tests/ping_test.c: Reorganize test, display error info.

        * src/lib/bootp_support.c: Add support for DNS options.

2000-02-29  Gary Thomas  <gthomas@cygnus.co.uk>

        * include/sys/sockio.h (SIOCSIFHWADDR): Add function to
        set hardware (MAC) address via ioctl.

2000-02-28  John Dallaway  <jld@cygnus.co.uk>

        * cdl/net.cdl:

        Reparent cdl_interface CYGHWR_NET_DRIVER_ETH0_SETUP under
        cdl_component CYGHWR_NET_DRIVER_ETH0_SETUP_OPTIONS to
        avoid a spurious conflict report when the latter is
        inactive.

2000-02-25  John Dallaway  <jld@cygnus.co.uk>

        * cdl/net.cdl:

        Reparent cdl_interface CYGHWR_NET_DRIVER_ETH1_SETUP under
        cdl_component CYGHWR_NET_DRIVER_ETH1_SETUP_OPTIONS to
        avoid a spurious conflict report when the latter is
        inactive.

2000-02-22  Gary Thomas  <gthomas@cygnus.co.uk>

        * include/netinet/if_ether.h: 
        * include/netinet/ip.h: 
        * include/netinet/ip_icmp.h: 
        * include/netinet/tcp.h: 
        * include/netinet/udp.h: More structure alignement fixes.

2000-02-21  Gary Thomas  <gthomas@cygnus.co.uk>

        * include/netinet/tcpip.h: 
        * include/netinet/udp_var.h: Force structure alignment.

2000-02-18  Gary Thomas  <gthomas@cygnus.co.uk>

        * tests/server_test.c: Target [client] address was wrong.

        * tests/tcp_source.c:
        * tests/tcp_sink.c:
        * tests/tcp_echo.c: New test suite.
        
2000-02-17  Gary Thomas  <gthomas@cygnus.co.uk>

        * tests/ping_test.c: Fix typo in startup message.

        * src/sys/net/route.c (route_reinit): New function - only here
        until BOOTP problems can be found/fixed.

        * src/lib/bootp_support.c (do_bootp): Force any existing routes
        to be reset since this causes a problem if BOOTP has been previously
        run.  Also, the BOOTP request must be zeroed or the server will
        get confused.

        * src/lib/network_support.c (init_all_network_interfaces): 
        Rearrange initialization so that all BOOTP activity takes place
        before any permanent setups are in place.

2000-02-16  Gary Thomas  <gthomas@cygnus.co.uk>

        * src/sys/net/route.c (rtioctl): Add support for deleting routes.

        * src/lib/network_support.c (init_all_network_interfaces): 
        Support predefined IP configurations, as well as BOOTP.

        * tests/ping_test.c: Renamed from 'bootp_test.c'

        * src/lib/inet_addr.c: 
        * include/arpa/ftp.h: 
        * include/arpa/nameser.h: 
        * include/arpa/telnet.h: 
        * include/arpa/inet.h: 
        * include/arpa/tftp.h: New file(s).

2000-02-15  Gary Thomas  <gthomas@cygnus.co.uk>

        * src/lib/select.c: Add 'cyg_select_with_abort()' interface which
        will allow 'select()' function to be abnormally terminated.  Add
        new function 'cyg_select_abort()' which does this.

2000-02-15  Hugo Tyson  <hmt@cygnus.co.uk>

        * cdl/net.cdl: Fix typo; cdl_interface CYGHWR_NET_DRIVER_ETH1 was
        missing, so EBSA build was conflictual.

2000-02-14  Gary Thomas  <gthomas@cygnus.co.uk>

        * tests/bootp_test.c: 
        * tests/ftp_test.c: 
        * tests/nc_test_framework.h: 
        * tests/nc_test_master.c: 
        * tests/nc_test_slave.c: 
        * tests/server_test.c: Rework with automatic network initialization.
        Includes support for multiple interfaces.

        * src/lib/network_support.c: 
        * include/network.h: New file(s).

        * cdl/net.cdl: New CDL to support multiple interfaces and their
        configuration.  (Not complete yet).

2000-02-11  Gary Thomas  <gthomas@cygnus.co.uk>

        * src/lib/select.c (select): Add a way to abort selects (EINTR).

2000-02-10  Gary Thomas  <gthomas@cygnus.co.uk>

        * tests/nc_test_slave.c (calibrate_load): Slightly different algorithm,
        which should work better on all platforms.

2000-02-09  Gary Thomas  <gthomas@cygnus.co.uk>

        * cdl/net.cdl: Add CYGPKG_NET_THREAD_PRIORITY to allow user more
        control over how networking [anonymous] threads behave.

        * tests/nc_test_framework.h: 
        * tests/nc_test_slave.c: 
        * tests/nc_test_master.c: Support test mode with target CPU under
        various loads.

2000-02-09  Hugo Tyson  <hmt@cygnus.co.uk>

        * src/ecos/support.c (bcmp): Fix bug.  Missing pointer increments
        meant that the zeroth byte only was compared.  Caused stack to
        reject ARP packets - depending on MAC address order for particular
        platform.

2000-02-09  Hugo Tyson  <hmt@cygnus.co.uk>

        * tests/bootp_test.c:
        * tests/ftp_test.c:
        * tests/nc_test_master.c:
        * tests/server_test.c:
        * tests/socket_test.c:
        Change stacksize of test thread to TYPICAL not MINIMAL - some
        drivers eat more stack than others.  Generally:
        -#define STACK_SIZE CYGNUM_HAL_STACK_SIZE_MINIMUM
        +#define STACK_SIZE CYGNUM_HAL_STACK_SIZE_TYPICAL

2000-02-08  John Dallaway  <jld@cygnus.co.uk>

        * cdl/net.cdl:

        Tidy display strings.

#####ECOSPDCOPYRIGHTBEGIN####
#
# Copyright (C) 2000, 2001, 2002 Red Hat, Inc.
# All Rights Reserved.
#
# Permission is granted to use, copy, modify and redistribute this
# file.
#
#####ECOSPDCOPYRIGHTEND####