OpenCores
URL https://opencores.org/ocsvn/openrisc/openrisc/trunk

Subversion Repositories openrisc

[/] [openrisc/] [trunk/] [rtos/] [ecos-3.0/] [packages/] [hal/] [mips/] [upd985xx/] [current/] [ChangeLog] - Rev 798

Go to most recent revision | Compare with Previous | Blame | View Log

2004-04-22  Jani Monoses <jani@iv.ro>

         * cdl/hal_mips_upd985xx.cdl :
         Invoke tail with stricter syntax that works in latest coreutils. 

2003-04-10  Nick Garnett  <nickg@balti.calivar.com>

        * src/hal_mips_upd985xx.ld: 
        Added libsupc++.a to GROUP() directive for GCC versions later than
        3.0.

2002-04-02  Hugo Tyson  <hmt@redhat.com>
2002-04-02  Anssi Pulkkinen <anssi.pulkkinen@ascom.com>

        * src/var_misc.c (cyg_hal_interrupt_acknowledge): Remove the read
        of the read-clear ISR register - it loses other pending interrupt
        sources.  Thanks to ASCOM for spotting this.  It should have been
        removed when the soft-copy and arbitration ISR were added, because
        those changes mean the hardware register would already be cleared
        down.

2002-01-22  Hugo Tyson  <hmt@redhat.com>

        * cdl/hal_mips_upd985xx.cdl: CYGIMP_HAL_COMMON_INTERRUPTS_CHAIN
        must not be enabled because the way the arbitration ISR is
        attached does not work with chained interrupts.

2001-10-30  Jonathan Larmour  <jlarmour@redhat.com>

        * cdl/hal_mips_upd985xx.cdl (CYGHWR_HAL_MIPS_UPD985XX_DIAG_BAUD): 
        This is the same as CYGNUM_HAL_VIRTUAL_VECTOR_CONSOLE_CHANNEL_BAUD
        which RedBoot uses, so define it.
        And also reimplement CYGINT_HAL_VIRTUAL_VECTOR_COMM_BAUD_SUPPORT
        as this is what makes it work.

2001-10-30  Hugo Tyson  <hmt@redhat.com>

        * cdl/hal_mips_upd985xx.cdl: Platform does *not* implement VV baud
        operations, in fact the implements
        CYGINT_HAL_VIRTUAL_VECTOR_COMM_BAUD_SUPPORT prevents it building.

2001-10-09  Hugo Tyson  <hmt@redhat.com>

        * cdl/hal_mips_upd985xx.cdl (..._MIPS_UPD985XX_HARDWARE_BUGS...):
        New CDL options to control workarounds for System Controller bugs
        S1 and S2.  Shorthand for S1 requires the fixes elsewhere.  That
        for S2 selects alternate versions of the interrupt code.

        * include/var_intr.h (HAL_INTERRUPT_ACKNOWLEDGE,...): Alternative
        versions which call the new routines in src/var_misc.c to deal
        with System Controller interrupts.

        * src/var_misc.c (cyg_hal_interrupt_unmask,...): New routines to
        manage the System Controller interrupts by software rather than
        S_ISR/S_IMR.  We never mask an interrupt once unmasked, but let
        the interrupt happen, fielding it silently.  When/if it becomes
        unmasked, then we call the ISR &c.

2001-08-17  Jonathan Larmour  <jlarmour@redhat.com>

        * cdl/hal_mips_upd985xx.cdl: Platform implements VV baud operations.

2001-08-08  Hugo Tyson  <hmt@redhat.com>

        * include/variant.inc (hal_intc_decode): Do not decode the S_ISR
        interrupts after all - they must be handled by arbitration.  So
        any S_ISR interrupt is reported as CYGNUM_HAL_INTERRUPT_SYSCTL -
        Who are they?  They are *all* number 6.

        * src/var_misc.c (_arbitration_isr): New routine to arbitrate
        between - and call all of - the interrupt sources that hang off
        the system controller S_ISR register, because the S_ISR register
        is read-clear, and the interrupt sources are edge-triggered so
        they do not re-assert themselves - so we must address multiple
        sources per actual interrupt.
        (hal_variant_init): Install _arbitration_isr() on the SYSCTL
        interrupt at startup.

        * include/var_intr.h: Commented the change in interrupt usage -
        the SYSCTL is occupied from time zero.

2001-08-01  Hugo Tyson  <hmt@redhat.com>

        * src/hal_mips_upd985xx.ld (hal_interrupt_sr_mask_shadow_base):
        This must be placed statically so it is shared with RedBoot.  And
        it must be an array so that we can address it from afar.

        * include/var_intr.h (HAL_DISABLE_INTERRUPTS): Implement this
        macro and its fellows as well as the individual MASK/UNMASK
        routines; these work together to keep the SR IM bits "true" to
        what is intended by means of a shadow variable.  Otherwise a race
        condition in the vanilla HAL_DISABLE_INTERRUPTS() can discard an
        interrupt mask change made by an ISR.
        (hal_interrupt_sr_mask_shadow): declare this and define it to
        entry zero of hal_interrupt_sr_mask_shadow_base[].
        
        * src/var_misc.c (hal_interrupt_sr_mask_shadow): initialize this
        new variable.

        * cdl/hal_mips_upd985xx.cdl (CYGPKG_HAL_MIPS_UPD985XX): set
        CYGINT_HAL_MIPS_INTERRUPT_RETURN_KEEP_SR_IM because we use bits
        within the SR for interrupt control.

2001-07-25  Hugo Tyson  <hmt@redhat.com>

        * include/var_intr.h (CYGNUM_HAL_INTERRUPT_SOFT_ZERO)
        (CYGNUM_HAL_INTERRUPT_SOFT_ONE): New interrupt numbers for
        software interrupts 0 and 1.  Also mask, unmask, and ack them
        correctly; ack clears the R/W bit in the cause regsiter.

        * include/variant.inc (hal_intc_decode) : Decode Software
        interrupts into numbers 0 and 1 - all else are moved up a bit.

2001-07-20  Hugo Tyson  <hmt@redhat.com>

        * cdl/hal_mips_upd985xx.cdl: Demand that CYGPKG_LIBM includes
        -fno-strict-aliasing in its CFLAGS_ADD to workaround a tools issue
        with access to double via pointer-to-union casts.

2001-07-18  Hugo Tyson  <hmt@redhat.com>

        * include/var_arch.h (UARTLCR_8N1): Add more divers definitions of
        UART control bits for implementing all the controls in the serial
        device that nobody ever uses.

2001-07-17  Hugo Tyson  <hmt@redhat.com>

        * include/variant.inc: If RAM startup, don't blow away the
        contents of cache - it might contain things that matter such as
        debug connection state.

2001-07-09  Hugo Tyson  <hmt@redhat.com>

        * include/var_arch.h (SDMDR_INIT): Change one of the numbers to
        match what's in the hardware when set up by customer code.

        * include/variant.inc: Remove dependency on temporary development
        config point CYGBLD_HAL_STARTUP_ROM_POST_OMIT_INITIALIZATION - the
        initialization now works OK.  So hal_memc_setup_table gets called.

        * src/variant.S (hal_memc_setup_table): ROM startup now works, so
        this code now gets called.

2001-07-06  Hugo Tyson  <hmt@redhat.com>

        * src/hal_diag.c (hal_uart_init): Ensure that we use the internal
        baud clock because there is no external one.

        * include/var_arch.h (UARTCLOCK): Clock is now 50MHz for "new"
        boards, not 18.xMHz any more.

2001-07-03  Hugo Tyson  <hmt@redhat.com>

        * include/var_intr.h (HAL_INTERRUPT_[UN]MASK): Better manipulate
        these with interrupts disabled for atomicity.

2001-07-02  Bart Veer  <bartv@redhat.com>

        * src/var_misc.c (hal_variant_init):
        Move the GPIO0 manipulating (flash programming voltagE)
        to the platform-init.

2001-06-27  Hugo Tyson  <hmt@redhat.com>

        This all fixes the serial interrupt problem of 2001-06-26.
        
        * src/var_misc.c (hal_variant_init): Initialize the system
        controller's ISR/IMR registers to mask and ack all external
        interrupts and then enable the underlying interrupt source that
        controls them all.
        Also provide hal_interrupt_handlers[CYGNUM_HAL_ISR_COUNT]; &c for
        use by system in vectors.S, and initialize them.

        * src/hal_diag.c (cyg_hal_plf_serial_isr): Do not paranoically
        acknowledge spurious interrupts.  No need.

        * include/variant.inc: Add complete redefinitions of interrupt
        decoding to handle extended interrupt sources from the system
        controller's ISR/IMR registers.

        * include/var_intr.h: Add complete redefinitions of interrupt
        management to handle extended interrupt sources from the system
        controller's ISR/IMR registers.

        * include/var_arch.h (S_ISR_ADR): Add ISR and IMR register
        addresses for use by assembly files.

        * include/plf_stub.h: Provide proto for cyg_hal_plf_comms_init()
        to reduce warning.

2001-06-26  Hugo Tyson  <hmt@redhat.com>

        * cdl/hal_mips_upd985xx.cdl: Remove src/var_stub.c - functionality
        is duplicated in the common HAL.

        * include/plf_stub.h: Remove a load of unneccessary cruft which is
        duplicated in the common HAL.

        * src/var_stub.c: Removed.

2001-06-26  Hugo Tyson  <hmt@redhat.com>

        * src/var_misc.c (hal_variant_init): Unmask the UART interrupt in
        the system controller - this is before the interrupt system in the
        MIPS core which is used for dynamic control.

        Also removed a load of duplicate hal_ctrlc_isr/HAL_CTRLC_ISR stuff
        that fought with the common HAL,

        Asynchronous CTRL-C still does not work.  The interrupt asserts
        forever.  This is similar to System Controller known bug S2
        "Interrupt Mask Restriction" from the NEC docs.  Hence this code
        in this file:   
        // *******************FIXME
        // This causes an interrupt loop from the UART as soon as you do IO.
        //    *S_IMR |= S_ISR_UARTIS; // unmask UART
        // *******************NB we mask it here so that the status is in control
        // *******************   of the application code whatever RedBoot does.
            *S_IMR &=~S_ISR_UARTIS;
        // *******************FIXME

        * src/hal_diag.c (cyg_hal_plf_serial_isr): This now contains all
        it should need to.
        (cyg_hal_plf_serial_control): Ditto, all cases now supported.

        CYGHWR_HAL_GDB_PORT_VECTOR use now unconditional, other tidyups.
        Bugfix to interrupt ack in nonblocking read, it was after the
        return.
        
        * cdl/hal_mips_upd985xx.cdl: We want to support ^Cm, so do not
        implement CYGINT_HAL_DEBUG_GDB_CTRLC_UNSUPPORTED.

        * include/var_intr.h (HAL_READ_INTR_REGS): New macro, handy for
        debugging MIPS.  Also remove duplicate hal_ctrlc_isr/HAL_CTRLC_ISR
        stuff that fought with the common HAL, just leaving
        CYGHWR_HAL_GDB_PORT_VECTOR defined.

        * src/var_stub.c (hal_var_stub_init): Use symbols for entries in
        vector tables and the like.

2001-06-22  Hugo Tyson  <hmt@redhat.com>

        * include/variant.inc: Use the proper target-agnostic config
        include CYGBLD_HAL_PLATFORM_H, and do not initialize the memory
        controller if you're told not to by the platform configuration
        CYGBLD_HAL_STARTUP_ROM_POST_OMIT_INITIALIZATION.

2001-06-22  Hugo Tyson  <hmt@redhat.com>

        * src/var_misc.c (hal_variant_init): Enable write-access to the
        flash area, and power-up the programming voltage via GPIO0, so
        that flash drivers can work.  Also enable the IBUS Arbiter so that
        internal peripherals can work.

        * src/variant.S (hal_memc_setup): Remove hal_memc_setup for RAM
        start; the work is done elsewhere.

        * include/variant.inc (hal_memc_init): Don't call hal_memc_setup
        for RAM start.
        
2001-06-07  Hugo Tyson  <hmt@redhat.com>

        * include/variant.inc (hal_memc_init): Always do the memc call
        even in RAM start, to enable...

        * src/variant.S (hal_memc_setup): Enable write-access to the flash
        area so that flash drivers can work.

2001-06-06  Hugo Tyson  <hmt@redhat.com>

        * include/var_cache.h (HAL_DCACHE_ENABLE_DEFINED): Add enough NOPs
        after diddling the cache-enability that it works.  Also give
        correct (apparantly!) figures for cache size despite the
        documentation arguing with itself.

        * src/var_misc.c (hal_variant_init): Enable the caches during
        startup.

        * include/var_arch.h: Add a very few definitions for use by
        assembler code here, and tidy up a little so that it can be used
        from .S files.  Specifically this is to let us init the RAM and
        ROM access controllers from assembly.

        * src/variant.S: 
        * include/variant.inc: Cut out some stuff we don't need that I had
        blindly copied from another platform.  Specifically we don't need
        to set up the TLB *at all* because all its space are belong, um,
        all memory, IO and devices are accessible through kseg1 and kseg0.
        Added initial (untested) cut at setup of RAM/ROM controllers.

2001-06-06  Hugo Tyson  <hmt@redhat.com>

        * cdl/hal_mips_upd985xx.cdl: Add implements statement for
        CYGINT_HAL_DEBUG_GDB_STUBS_BREAK, moved from the platform HAL.

2001-06-05  Hugo Tyson  <hmt@redhat.com>

        * include/variant.inc: It all works rather better now...  fiddling
        with vector setup and cache initialization.

2001-06-05  Hugo Tyson  <hmt@redhat.com>

        * cdl/hal_mips_upd985xx.cdl: to make GDB stubs work, implements
        CYGINT_HAL_MIPS_STUB_REPRESENT_32BIT_AS_64BIT.  Commented out
        implements CYGINT_HAL_VIRTUAL_VECTOR_SUPPORT_NOT_GUARANTEED.

        * include/hal_diag.h: Now it works, properly be dependent on
        CYGSEM_HAL_VIRTUAL_VECTOR_DIAG.

        * include/plf_stub.h: Choose the right init routine dependent on
        CYGSEM_HAL_VIRTUAL_VECTOR_DIAG.

        * src/hal_diag.c (hal_uart_init): Remove diagnostics and delays
        from initialization.

        * src/var_stub.c (hal_var_stub_init): Remove bogus definition of
        vsr_table and leave breakpoint VSR alone, the springboard will
        handle it AOK.

2001-06-04  Hugo Tyson  <hmt@redhat.com>

        * cdl/hal_mips_upd985xx.cdl
        * include/hal_diag.h
        * include/plf_stub.h
        * include/var_arch.h
        * include/var_cache.h
        * include/var_intr.h
        * include/variant.inc
        * src/hal_diag.c
        * src/hal_mips_upd985xx.ld
        * src/var_misc.c
        * src/var_stub.c
        * src/variant.S
        New files; initial checkin.

        RAM startup works; kernel tests run, loaded via SRecords,
        including tm_basic, clockcnv, mutex3 and the exception tests.
        Virtual vector calling works, but all apps do not work with a ROM
        monitor, so they are all standalone.  Diag printf et all all work.
        Floating point emulation works.

//===========================================================================
// ####GPLCOPYRIGHTBEGIN####                                                
// -------------------------------------------                              
// This file is part of eCos, the Embedded Configurable Operating System.   
// Copyright (C) 1998, 1999, 2000, 2001, 2002 Free Software Foundation, Inc.
//
// This program 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.                                                           
//
// This program 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 this program; if not, write to the                            
// Free Software Foundation, Inc., 51 Franklin Street,                      
// Fifth Floor, Boston, MA  02110-1301, USA.                                
// -------------------------------------------                              
// ####GPLCOPYRIGHTEND####                                                  
//===========================================================================

Go to most recent revision | Compare with Previous | Blame | View Log

powered by: WebSVN 2.1.0

© copyright 1999-2024 OpenCores.org, equivalent to Oliscience, all rights reserved. OpenCores®, registered trademark.