Subversion Repositories openrisc_me

2003-02-12  Mark Salter  

        * src/if_i82559.c: Remove restriction preventing combination of

        *_GET_ESA and *_HAS_EEPROM.

2003-01-03  Gary Thomas  

        * src/if_i82559.c:

        * cdl/intel_i82559_eth_drivers.cdl: Allow finer control over

        debug (chatter) by making control booldata.

2002-12-17  Mark Salter  

        * src/if_i82559.c (CYGHWR_DEVS_ETH_INTEL_I82559_ENDIAN_NEUTRAL_IO): New

        flag to support systems where PCI IO operations are not affected by

        CPU endianess.

2002-10-06  Andrew Lunn  

        * src/if_i82559.c (pci_init_find_82559s): Changed scope of

        max_interrupt_handle since its needed in other places.

        * src/if_i82559.c (i82559_stop): Corrected wrong name of variable

        in a diag_printf function.

2002-07-24  Gary Thomas  

        * src/if_i82559.c (pci_init_find_82559s):

        Allow platform to define CYGHWR_DEVS_ETH_INTEL_I82559_USE_MEMORY

        which forces use of memory instead of I/O space for device access.

2002-06-14  Gary Thomas  

        * src/if_i82559.c:

        Need to include  for proper configuration

        of stand-alone (polled) vs. system (interrupt driven) mode.

2002-04-09  Hugo Tyson  

        * src/if_i82559.c (i82559_stop): Free up any pending transmissions

        to prevent MBUF store leaks if the calling thread - the one that's

        shutting down the interface - priority ends up higher than (or

        same as) the network service threads.

2002-03-26  Jonathan Larmour  

        * src/if_i82559.c (wait_for_cmd_done): Don't assert if RUC_ADDR_LOAD

        command doesn't clear - on some chips it never does.

2002-03-18  Hugo Tyson  

        * src/if_i82559.c (pci_init_find_82559s): Move allocation of

        statistics pointer for each device to here, out of

        i82559_start().  It used to leak PCI window store every time the

        interface cycled down/up.  [Case 107110]

2002-02-25  Mark Salter  

        * src/if_i82559.c: Don't call PacketRxReady in standalone configuration

        unless ready to accept a packet. Limit packet delivery to one packet at

        a time in standalone configuration.

2002-02-20  Hugo Tyson  

        * include/i82559_info.h: New fields to record/control multicast

        all reception, oversized packet reception and promiscuous mode

        neatly.

        * src/if_i82559.c: Initial support for multicast address reception

        filtering.

        (i82559_restart): New routine to restart the device after fiddling

        with its setup - a tail recursion from i82559_start().

        (i82559_configure): More generic flags passed in and handled.

        (i82559_set_multicast): Routine to pass a set of addresses into

        the device; it works out the bitmap et al for us!

        (i82559_ioctl): Handle new case arms ETH_DRV_SET_MC_LIST and

        ETH_DRV_SET_MC_ALL if defined.  Reset and restart the device,

        diddling its status between.

2002-02-19  Mark Salter  

        * src/if_i82559.c (pci_init_find_82559s): Don't use IRQ support if

        CYGPKG_IO_ETH_DRIVERS_STAND_ALONE.

2002-01-28  Mark Salter  

        * src/if_i82559.c (pci_init_find_82559s): Removed CYG_ASSERT which

        assumed system PCI window exists.

2002-01-10  Jonathan Larmour  

        * src/if_i82559.c (find_82559s_match_func): Add some more compatible PCI

        ids.

2002-01-03  Mark Salter  

        * src/if_i82559.c (pci_init_find_82559s): Remove check that

        assumes driver owns entire PCI window.

        * cdl/intel_i82559_eth_drivers.cdl: Add

        CYGNUM_DEVS_ETH_INTEL_I82559_MAX_TX_DESCRIPTORS and

        CYGNUM_DEVS_ETH_INTEL_I82559_MAX_RX_DESCRIPTORS.

2001-11-19  Hugo Tyson  

2001-11-19  Anssi Pulkkinen 

        * src/if_i82559.c (TxMachine): The test to see if there is a new

        tx to start must also say "go" if the tx queue is full.  Normally,

        if the txqueue is full, a tx will be occurring at all times - so

        one will complete soon, so the tx queue full flag will soon be

        cleared, and this condition will recover.  I suspect a subtle race

        which effectively means a new tx is queued and fills the queue in

        between two tx's, so no TxDone() follows, causes the hang which we

        get without this extra test under high load.  [CASE 106686]

2001-09-20  Gary Thomas  

        * src/if_i82559.c: Add declaration of 'i82559_stop' to quiet warning.

2001-09-13  Hugo Tyson  

        * cdl/intel_i82559_eth_drivers.cdl:

        * src/if_i82559.c:

        Remove the tickle thread from this driver.  The functionality is

        now in the network timeout functionality which handles the

        "deliver" calls.  So it's only called if the net is quiet.

2001-09-12  Hugo Tyson  

        * src/if_i82559.c (TxDone): Defensive programming; zero the Tx key

        in the global data before making the callback.  This change

        appeared useful in our forebear, the EBSA-specific driver.

2001-08-29  Hugo Tyson  

        * cdl/intel_i82559_eth_drivers.cdl: New CDL options called

        CYGPKG_DEVS_ETH_INTEL_I82559_TICKLE_THREAD, .._PRIORITY and

        .._DELAY to control the thread.  By default this is all disabled

        because it is not needed for a real network application which

        itself proactively tries to use the net - it's only needed when

        the app is totally passive, so the driver otherwise does not get

        CPU time at all.

        * src/if_i82559.c (starti82559ticklethread): New code to start a

        thread (!) to catch lost interrupts and thus restart jammed

        hardware.  It's only possible if KERNEL and NET - RedBoot is

        unaffected, and it's not needed for RedBoot 'cos RedBoot polls.

        (TxMachine): An assert about the CU idle status sometimes fired;

        made the code more conservative wrt timing here.

        (comments): more documentation on the possible external configury

        by .inl file, described the two TIMEOUT macros that can be set,

        and indeed must be to use the tickling thread meaningfully.

2001-08-22  Gary Thomas  

        * src/if_i82559.c:

        printf() is no longer a part of RedBoot.  Thus all programs

        must use diag_printf() and related functions instead.

2001-06-22  Jesper Skov  

        * src/if_i82559.c (i82559_init): Handle EEPROMS without CRC.

2001-05-16  Jonathan Larmour  

        * src/if_i82559.c (udelay): Use virtual vector rather than assuming

        hal_delay_us() exists.

2001-04-09  Hugo Tyson  

        * src/if_i82559.c (pciwindow_mem_alloc): If asked, by

        CYGHWR_DEVS_ETH_INTEL_I82559_PCIMEM_DISCONTIGUOUS, check for

        breaks in the physical address of PCI window memory, and discard

        any block with a break in it, then redo the alloc.  This is for

        targets where smaller SDRAM chips are fitted than the address

        width, so they end up with gaps in (physical) memory.

2001-04-09  Hugo Tyson  

        * src/if_i82559.c (i82559_init): Support hardware with only one

        serial EEPROM attached to one device, but multiple devices.  This

        generally involves moving some decisions for eg. programming

        EEPROM or not, optionally from compiletime to runtime.  Improved

        the diagnostics a bit also, and fixed some bugs.

        (eth_set_mac_address): now takes an arg to say whether we want to

        program the EEPROM, or are just setting up the chip.

        (i82559_ioctl): This is the only place we set that arg to write

        the EEPROM, startup never tries to write EEPROM from any external

        source such as configured ESA or RedBoot params.

2001-04-09  Hugo Tyson  

        * src/if_i82559.c (i82559_start): Warnings fixed; unused vars

        following previous change.

2001-04-09  Hugo Tyson  

        * src/if_i82559.c (i82559_start): Do not wait for command

        completion after the initial DUMPSTATS operation.  The code to

        wait was bogus; it was looking at the structure in the wrong way.

        In any case, there is no need to wait, the wait_for_cmd_done() in

        any following activity is good enough.

2001-03-28  Hugo Tyson  

        * src/if_i82559.c (update_statistics): Fix warning - only declare

        op_i82559 if we use it - from previous change.

2001-03-23  Hugo Tyson  

        * src/if_i82559.c (i82559_can_send, i82559_send): Count up and

        down entries into these routines, and always say "no" to

        can_send() if we are recursing, ie. within_send != 0.  This is a

        good idea because it limits stack usage.  The recursion happens

        because of the very neccessary polling of rx state within the tx

        routines - which can deliver packets up into the stack, which can

        trigger a response tx and so on...  Also fixed a harmless bug

        where, after polling, the p_i82559 current device state pointer

        pointer to the wrong device, if we polled both devs regardless; it

        only caused problems with the within_send recursion count.

        * include/i82559_info.h (I82559): Add within_send field.

2001-03-15  Hugo Tyson  

        * src/if_i82559.c (i82559_init): Support external macro

        CYGHWR_DEVS_ETH_INTEL_I82559_GET_ESA(...), which if defined, is

        used to acquire a MAC address, on the assumption that there is no

        EEPROM present.  CYGHWR_DEVS_ETH_INTEL_I82559_HAS_NO_EEPROM

        confirms that there is no EEPROM attached to the 82559(s) so we

        can omit all that code.  It's in the negative sense because it's

        the unusual case.

2001-03-13  Hugo Tyson  

        * src/if_i82559.c (CYGHWR_DEVS_ETH_INTEL_I82559_DEMUX_ALL): Fully

        enable "DEMUX_ALL" option; all devices are scanned for

        activity/wedging/interrupts at every opportunity.  Multiple

        devices on the same interrupts are now supported.

2001-03-12  Hugo Tyson  

        * src/if_i82559.c (i82559_configure): New routine combines actions

        of eth_set_promiscuous_mode() and eth_set_config() both of which

        disappear.  Thus the device is always configured correctly from

        the start.  [from a patch from Martin Buck  

        via the EBSA285 driver]

        (eth_dsr): Do not call upper layer unless we are in a networked

        [non-RedBoot, non-polled] environment.  This is necessary because

        we call our own DSR from foreground to unstick after a lost

        interrupt.

        (find_82559s_match_func): New routine to discover disparate

        devices on the PCI bus.  Used in pci_init_find_82559s().

        (dump_txcb): Various junk debug functions removed.

2001-03-12  Hugo Tyson  

        * src/if_i82559.c: Some tidying up, and two main additions to the

        mechanisms used for managing this nasty entomological device.

        (i82559_can_send): [amongst other places] if it is defined, use

        CYGHWR_DEVS_ETH_INTEL_I82559_MISSED_INTERRUPT(p_i82559) to detect

        that we missed an interrupt - and if so, call the ISR and DSR

        directly.  This is necessary for edge-triggered interrupt

        controllers being fed by this level-sending device, where an

        internal event in the device can keep the line low after the code

        thinks that all events have been handled.

        (Check82559TxLockupTimeout): This routine runs a timeout (if the

        HAL provides a pair of watchdog-like macros) which performs a

        selective reset of the device if it takes too long to transmit.

        This happens rarely, but when it happens this is the only way out;

        the tx machine is fixated on one tx and never comes out of it.

        Sadly these two techniques to escape lockup only work if the

        driver is called (ie. the stack tries to tx); incoming packets

        cannot unwedge a device.  We might need to add a poll from the

        stack in future if this causes problems.

        * include/i82559_info.h (I82559): Added two words for managing tx

        lockup timeouts since this must be per-device.

2001-03-02  Hugo Tyson  

        * src/if_i82559.c: Remove "Platform specific - defaults provided"

        definitions for CYGHWR_INTEL_I82559_PCI_MEM_MAP_BASE/SIZE; all

        platforms now provide these.

        Added a little further disclaimer to the comment about LE/BE/GE.

2001-03-01  Hugo Tyson  

        * src/if_i82559.c: Global change: what was HAL_READ_UINTxx or

        HAL_WRITE_UINTxx are now READMEMxx or WRITEMEMxx - just

        convenience macros that do all the volatile casts we want there.

        Definitions of INL, INW, INB and OUTL, OUTW, OUTB recast in terms

        of HAL_READ_UINTxx or HAL_WRITE_UINTxx so that they will use

        proper IO operations on those CISCy mainframes that have a special

        way of doing IO.

2001-03-01  Hugo Tyson  

        * src/if_i82559.c: Much more configury added, so that I can use it

        with a new target board.  Including, but not limited to:

        o An essay about addressing, big endian, little endian and GIB

        endian (sic) and how we treat the various types of access within

        this module.  In other words, a lot of comments.

        o CYGNUM_DEVS_ETH_INTEL_I82559_SEPARATE_MUX_INTERRUPT to clarify

        that such an interrupt is SEPARATE from any simplex intrs that are

        also available.

        o CYGHWR_DEVS_ETH_INTEL_I82559_DEMUX_ALL for hardware where the

        the only interrupt is a multiplexed (wire or'd) one.  Plus

        associated macros for direct manipulation of interrupt masks and

        acknowledgments.

        o Support for external definition from the platform HAL or driver

        config module of CYGHWR_INTEL_I82559_PCI_MEM_MAP_BASE and SIZE.

        Also CYGHWR_INTEL_I82559_PCI_VIRT_TO_BUS.

        o LE definitions of a few macros that were missing.

        o Separated definitions of structure offsets from BE/LE

        definitions of consts within the words.  Offsets vary with GIB

        endian, for those CPUs who just flip bits.  No change for BE

        machines, needed for LE machines to work.  To enable address-flip

        within structures, define CYG_ADDRESSING_IS_GIBENDIAN.

        o Re-org of mux/simplex deliver and isr's for new configuration.

        o Couple of extra debug printouts.

2001-01-26  Jesper Skov  

        * src/if_i82559.c: Added FIXME for promiscuous mode.

2001-01-25  Jesper Skov  

        * src/if_i82559.c: Replace CYGNUM_HAL_INTERRUPT_PCI_IRQ with

        CYGNUM_DEVS_ETH_INTEL_I82559_MUX_INTERRUPT which platform .inl

        must provide when appropriate. Changed _deliver functions to match

        _isr functions.

        * include/i82559_info.h: Remove debug hacks.

        Fix declaration.

        * src/if_i82559.c: Moved device descriptors into platform header.

        Allow individual devices to hardwire ESA. Handle N devices instead

        of just 1 or 2.

        * include/i82559_info.h: ESA hardwired flag added.

        * cdl/intel_i82559_eth_drivers.cdl: Device details moved to

        platform CDL.

        * src/if_i82559.c (i82559_start): Poll status after dump command.

2001-01-24  Jesper Skov  

        * src/if_i82559.c: [changes from yesterday] Some more endian

        fiddle, and the device coughs out the initial ARP packets.

        (wait_for_cmd_done): semantics changed to wait for pending CU

        cmds.

        (i82559_reset): Cleaned up to spec. Init register bases after

        reset.

        Fix initialization of hardwired ESA.

        Correct configuration command.

        Fix rcv len masking.

        (eth_set_mac_address): Set driver's ESA as well. Proper check for

        completion.

        Fix LE structure offsets.

        (i82559_start): Call new eth_set_config to enable device. Last

        hacks cleaned up.

        Remove a few printfs.

2001-01-23  Jesper Skov  

        * src/if_i82559.c: Debug hackery and some endian issues resolved.

2001-01-22  Jesper Skov  

        * src/if_i82559.c: Major changes, getting rid of device structures

        in favor of HAL IO accessor macros. Also added endian conversion

        macros where required.

        * include/i82559_info.h: Change type of device structures to char.

2001-01-19  Jesper Skov  

        * cdl/intel_i82559_eth_drivers.cdl: Hack for just one interface

        now.

        * src/if_i82559.c: Fix IO functions. Handle EEPROM not being

        attached to device.

        * src/if_i82559.c: Use uncached/physical address translation

        macros.

2001-01-15  Jesper Skov  

        * src/if_i82559.c: Handle endian differences between controller

        and CPU. Leave usdelay to HAL. Hack around PCI-base-at-0

        assumption.

        Increase SK_DELAY (not that it changed anything, but now it's to

        the spec), hardwire static ESA.

2001-01-12  Jesper Skov  

        * src/if_i82559.c (pci_init_find_82559s): Check that device has

        been found before accessing it.

        (pci_init_find_82559s): Only delay if devices were

        found. Recognize 82559ER code as well.

2001-01-10  Jesper Skov  

        * src/if_i82559.c: Minor hacks to get it to build.

        * include/i82559_info.h: Same.

        * cdl/intel_i82559_eth_drivers.cdl: Same.

2000-12-07  Jesper Skov  

        * cdl/intel_i82559_eth_drivers.cdl: Cloned from the EBSA driver.

        * include/i82559_info.h: Same.

        * src/if_i82559.c: Same.

2000-11-19  Gary Thomas  

        * src/if_ebsa285.c (pci_init_find_82559s): Intel has at least

        two devices equivalent to the 82559.  Support both (0x1229, 0x01030).

2000-10-05  Hugo Tyson  

        * src/if_ebsa285.c: Deal with device interrupts in a nested

        fashion - disable/restore is the semantics now, rather than

        unconditionally unmasking.  Also go directly to the 21285 PIC's

        interrupt control registers to gain atomicity for these.  Poll for

        ready received packets when acknowledging an interrupt in the

        tranmitting world; a race here could lose an Rx interrupt.  Which

        doesn't matter on a busy system, but in quieter times...  there

        will always be such a race because of the vague way the i82559's

        status bits reflect how it's yanking the interrupt line; you have

        to poll until the interrupt is gone before returning else spurious

        interrupt failures occur.  The issue is to close the window as

        tightly as possible, which this change achieves at a minor cost in

        performance - because of the time spent polling when not required.

2000-09-11  Hugo Tyson  

        * src/if_ebsa285.c (i82559_poll): Only diddle the interface we

        were asked to.  This is more correct in terms of the intent of the

        API, though it shouldn't really be necessary.

2000-09-06  Hugo Tyson  

        * src/if_ebsa285.c (pci_init_find_82559s): Add asserts and an

        unconditional check that the PCI window as configured matches the

        address and size of the pci_window region from the MLT.  This is

        here because at present the MLT and CT cannot intercommunicate

        enough.  The separation of the PCI window is needed because

        otherwise the malloc heap will eat all memory.  [This is related

        to CR 902624-CR, "MLT needs to be configuration aware"]

2000-09-01  Hugo Tyson  

        * 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/if_ebsa285.c: Minor: fix the big assert in i82559_send()

        which suffered a race condition when called from the fast thread

        rather than from a DSR.  Major: Add a "deliver" entry to the

        interface record for the "fast thread" implementation of the

        network internal comms system.  Provide a pass-up DSR to the

        logical ether driver's DSR and appropriate delivery routine(s).

        i82559_poll() now calls i82559_deliver() rather than the DSR.  Add

        valid data for mux'd DSR to pass on up.

2000-09-01  Hugo Tyson  

        * tests/test_net_realtime.h: Tighten up the latency requirements

        by a factor of 5; it all seems happy, so committed.

2000-08-25  Hugo Tyson  

        * src/if_ebsa285.c (i82559_ioctl): A little further diddling; have

        a bool to say whether the dot3 info is filled in.

2000-08-24  Hugo Tyson  

        * cdl/ebsa285_eth_drivers.cdl: Do not export a symbol for the

        device info file (include/ebsa285_info.h) since nobody needs

        (portably) to include it now.

        * src/if_ebsa285.c (i82559_ioctl): Handle new ioctl calls

        ETH_DRV_GET_IF_STATS_UD and ETH_DRV_GET_IF_STATS to get loads of

        statistical information out.  _UD means update.  The nonUD one can

        be used instead mostly, if we find the performance hit too large.

        This should allow SNMP (a) to not explode, (b) to get useful info

        out of other device implementations than this one.

        * include/ebsa285_info.h: Remove all the macro cruft for feature

        detecting of lots of individual statistics; we now just have a

        catch-all struct that SNMP uses, defined in the common ether

        driver environment.

2000-08-15  Hugo Tyson  

        * src/if_ebsa285.c (PacketRxReady): Put back the check for very

        small packets into the driver; the layer above checks for that

        (defensive programming) but only *after* asserting that the size

        is large enough, to help detect that scenario from other drivers.

        I believe we only have struct ether_header available if CYGPKG_NET

        but I could be wrong.

        [CASE 104353]

2000-08-08  Hugo Tyson  

        * src/if_ebsa285.c (eth_set_promiscuous_mode):

        - ccs->config_bytes[18]=0x70;

        + ccs->config_bytes[18]=0x72; // Keep the Padding Enable bit

        ...otherwise short frame sends don't work in promisc mode.

        [CASE 104289]

2000-08-07  Gary Thomas  

        * src/if_ebsa285.c (pciwindow_mem_alloc): Take out very noisy debug.

2000-08-03  Gary Thomas  

        * src/if_ebsa285.c: Changes for stand-alone mode.

        * cdl/ebsa285_eth_drivers.cdl: Ethernet driver package hierarchy changed.

        Add option to control number of interfaces.

2000-07-28  Hugo Tyson  

        * src/if_ebsa285.c (PacketRxReady): Do not attempt to forward

        short packets; eth_drv.c assumes there is at least a header there.

        (i82559_recv): Also be more careful and ASSERTive about -ve buffer

        sizes; be more defensive about sglists.  [CASE 104206]

2000-07-26  Gary Thomas  

        * src/if_ebsa285.c: Update for new eth_drv interfaces.

2000-07-18  Hugo Tyson  

        * src/if_ebsa285.c (i82559_poll): Fill in the flesh of this, it

        just calls ISR and DSR repeatedly.

        (i82559_start): Look in the device record for promiscuous mode

        flag; it should be passed though the common layer, but it's not

        [yet] - this change from Andrew Lunn/ASCOM.  Also a fix and delay

        to the promisc mode code per se.

2000-07-17  Hugo Tyson  

        * src/if_ebsa285.c (i82559_poll): New function, just to fill in

        the interface record; not used.

2000-06-27  Hugo Tyson  

        * cdl/ebsa285_eth_drivers.cdl: Add sesquipedalian option

        CYGDBG_DEVS_ETH_ARM_EBSA285_KEEP_82559_STATISTICS in (now)

        component CYGDBG_DEVS_ETH_ARM_EBSA285_KEEP_STATISTICS to control

        keeping (well, harvesting really) the i82559's internal stats.

        Reputedly, it doesn't service the net whilst this is happening, so

        it could be viewed a bad thing.  Hence the option.

        * include/ebsa285_info.h: Only describe the I82559_COUNTERS

        i82559_counters[2]; structs if full stats are to be kept.

        * src/if_ebsa285.c (update_statistics): Only include this if full

        stats are to be kept.

2000-06-27  Hugo Tyson  

        * src/if_ebsa285.c (ResetRxRing): Re-do the management of the

        RxRing; have an end-of-list flag (EL) in the last entry, and as

        you unload filled slots, drag it round after you.

2000-06-14  Hugo Tyson  

        * cdl/ebsa285_eth_drivers.cdl: Add option to control statistics

        keeping.

        * include/ebsa285_info.h: Act on it.

2000-06-13  Hugo Tyson  

        * cdl/ebsa285_eth_drivers.cdl: Properly include the new header

        file and define symbols to allow client code to get at it without

        knowing the particular device driver name.

        * include/ebsa285_info.h: New file: export various statistics

        information about the driver for use by monitoring and

        network-management systems.  This requires exposing the

        (otherwise) internal structures of the driver.

        * src/if_ebsa285.c: remove a lot of structure definitions &c that

        are now necessarily in the new header; add a couple of new

        routines which provide status and update statistics from the

        device into memory; tidy up control of whether stats-keeping is

        enabled.

2000-06-06  Hugo Tyson  

        * cdl/ebsa285_eth_drivers.cdl (define_proc): Add #define of

        CYGDAT_DEVS_ETH_DESCRIPTION in the config file for information.

2000-05-12  Hugo Tyson  

        * tests/test_net_realtime.h (tnr_print_activity): New routine to

        check the system is working, tidied up the API.  It works!

2000-05-11  Hugo Tyson  

        * cdl/ebsa285_eth_drivers.cdl: Added export of the test header

        below, and config opts for controlling EEPROM writing and all the

        status chatter as the device starts up.

        * src/if_ebsa285.c: Reworked the code for reading and setting the

        EEPROM that holds the MAC address.  This is very ugly, but now

        more reliable.  Also tidied up printing cruft with neater

        configury, and made it an option (for safety) whether it's

        possible to write the EEPROM at all.

        * tests/test_net_realtime.h: New file - it is intended to be used

        by networking tests to verify that latency is not compromised by

        the stack and driver.  It's very platform specific, hence the

        location in here.  This is a preliminary version only.

2000-04-27  Hugo Tyson  

        * src/if_ebsa285.c: A serious re-write.  This cuts out a lot of

        code from the old version and improves the performance greatly.

        The cruft was mainly doing lots of explicit event communication

        between the ISR and DSR, when in fact all the state needed is

        present in the tx/rx rings.  So both ISRs and DSRs regard their

        call as an opportunity to progress everything they can, rather

        than only dealing with one interrupt cause at a time; the

        connection between them is now rather looser.

        Interrups can now be re-enabled after the ISR (in other words they

        are not masked in the ISR), no need to wait for the DSR, but in

        consequence some DSR code must mask/unmask intrs as it works.

        The 82559 appears to be a little slow in reacting to commands and

        state changes, so some interrupts were being lost - or persisting

        beyond their desired life - so there's some kinda polling code to

        deal with that also.  We also rely on the foreground to kind of

        poll in the same way, in the send/can_send calls - we know the

        stack will re-try if necessary, though this is rare.

        The driver now works (albeit at much reduced performance) with as

        few as 6 rx and tx buffers - in other words the "queue full/out of

        rx buffers" states have been tested and all is well.  It works

        generally fine with 8 buffers of each kind.

        The mux ISR and DSR are now rather more polled than the old

        versions; we just try to do things with both devices (if active)

        by simply calling each unitary ISR/DSR respectively.

        I also re-ordered some of the code, moving utilities to the end of

        the file and grouping together Tx and Rx machines a bit better.

2000-04-13  Hugo Tyson  

        * src/if_ebsa285.c: Attribution to Ron Spence, Pacific Softworks

        added as a contributor.

2000-04-07  Hugo Tyson  

        * ecos.db: Re-organize device packages.  This is a massive change

        involving deleting all the sources for serial and ethernet drivers

        from where they used to live in

            packages/io/serial/current/src/ARCH/PLATFORM.[ch]

            packages/net/drivers/eth/PLATFORM/current/src/...

        and reinstating them in

            packages/devs/serial/ARCH/PLATFORM/current/src/...

            packages/devs/eth/ARCH/PLATFORM/current/src/...

        All these new packages are properly defined in ecos.db, and are

        all of type "hardware" so that a "target" can grab them.

        This directory layout is descriptive of the devices we have right

        now, arch and platform are separate levels just to make it easier

        to navigate in the filesystem and similar to the HAL structure in

        the filesystem.

        It is *not* prescriptive of future work; for example, the mythical

        common highly-portable 16550 serial driver which works on many

        targets would be called "devs/serial/s16550/current", or a serial

        device for a particular board (cogent springs to mind) that can

        work with different CPUs fitted is "devs/serial/cogent/current".

        Changelogs have been preserved and replicated over all the new

        packages, so that no history is lost.

        The contents of individual source files are unchanged; they build

        in just the same emvironment except for a very few cases where the

        config file name changed in this movement.

        Targets in ecos.db have been redefined to bring in all relevant

        hardware packages including net and serial drivers (but the newly

        included packages are only active if their desired parent is

        available.)

        The names of CDL options (and their #defines of course) stay the

        same for the serial drivers, for backward compatibility.

        * templates/*/current.ect: these have had CYGPKG_IO_SERIAL added

        rather than it being in (almost) all target definitions.

2000-03-29  Hugo Tyson  

        * src/if_ebsa285.c (i82559_recv): Be happy with NULLs in the

        SGlist; it means the caller is out of memory so drop the packet on

        the floor.  Also remove a completely redundant test.

2000-03-06  Hugo Tyson  

        * src/if_ebsa285.c (i82559_can_send): Update net driver to new

        interface style.  This is incomplete wrt promiscuous mode, but

        that's probably about all.

2000-02-14  Gary Thomas  

        * cdl/ebsa285_eth_drivers.cdl: Update CDL to indicate multiple

        interface support.

2000-02-14  Hugo Tyson  

        * src/if_ebsa285.c (pci_init_find_82559s): Tidy comments somewhat

        and set debug and stats collecting defines to most friendly

        settings.

2000-02-10  Hugo Tyson  

        * src/if_ebsa285.c (PacketRxReady): Fix bug; current descriptor

        was not being write back for the callback to use.  Hence asserts

        on state of rfd were firing in busy times - that leading rfd had

        already been drained.

        Also rationalized meaning of DEBUG printy symbols a bit - it's now

        chatty during startup/foreground manipulations but not in any

        performance related activities ie. rx or tx.

2000-02-09  John Dallaway  

        * cdl/ebsa285_eth_drivers.cdl:

        Reparent under CYGPKG_NET_ETH_DRIVERS and tidy display strings.

2000-02-08  Hugo Tyson  

        * src/if_ebsa285.c: New File.

        * cdl/ebsa285_eth_drivers.cdl: New File.

        Initial Checkin of EBSA285 Ethernet driver.

        It's one monolithic file at present, and should be split up into a

        more generic Intel 82559 driver plus platform-specific parts (PCI

        et al) plus eCos/Red-Hat-BSD-stack parts.

//===========================================================================

//####ECOSGPLCOPYRIGHTBEGIN####

// -------------------------------------------

// This file is part of eCos, the Embedded Configurable Operating System.

// Copyright (C) 1998, 1999, 2000, 2001, 2002 Red Hat, Inc.

//

// eCos is free software; you can redistribute it and/or modify it under

// the terms of the GNU General Public License as published by the Free

// Software Foundation; either version 2 or (at your option) any later version.

//

// eCos is distributed in the hope that it will be useful, but WITHOUT ANY

// WARRANTY; without even the implied warranty of MERCHANTABILITY or

// FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License

// for more details.

//

// You should have received a copy of the GNU General Public License along

// with eCos; if not, write to the Free Software Foundation, Inc.,

// 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA.

//

// As a special exception, if other files instantiate templates or use macros

// or inline functions from this file, or you compile this file and link it

// with other works to produce a work based on this file, this file does not

// by itself cause the resulting work to be covered by the GNU General Public

// License. However the source code for this file must still be made available

// in accordance with section (3) of the GNU General Public License.

//

// This exception does not invalidate any other reasons why a work based on

// this file might be covered by the GNU General Public License.

//

// Alternative licenses for eCos may be arranged by contacting Red Hat, Inc.

// at http://sources.redhat.com/ecos/ecos-license/

// -------------------------------------------

//####ECOSGPLCOPYRIGHTEND####

//===========================================================================