From mboxrd@z Thu Jan 1 00:00:00 1970 From: Gary Thomas To: Jamie Guinan Cc: ecos-discuss@sourceware.cygnus.com, Jonathan Larmour Subject: RE: [ECOS] MBX stubrom (was [ECOS] loader.bin sources?) Date: Wed, 21 Jul 1999 07:06:00 -0000 Message-id: References: X-SW-Source: 1999-07/msg00026.html Here's the script I use to do this (running on another platform). Perhaps it will lend some help. (Note: we're working on improving both the documentation and procedure for making stubs...) ----------------- STUBS_config ------------------------------------ #! /bin/sh tcl $1/pkgconf.tcl \ --target=arm --platform=pid --startup=stubs \ --disable CYGPKG_KERNEL --disable CYGPKG_UITRON \ --disable CYGPKG_LIBC --disable CYGPKG_LIBM \ --disable CYGPKG_ERROR --disable CYGPKG_IO \ --disable CYGPKG_IO_SERIAL --disable CYGPKG_DEVICES_WALLCLOCK \ --disable CYGPKG_DEVICES_WATCHDOG patch < On Sun, 18 Jul 1999, Jamie Guinan wrote: > >> What I want to do next is make an stubrom image for the FADS 823 >> board. I understand that the FADS is unsupported, but my client >> (he has a Cygnus support contract) pointed me at the MBX HAL >> for an example. For starters, it looks like the Cogent has >> external 16550 uarts whereas the MBX and FADS use on-chip >> SMCs or SCCs. >> >> Bummer, I didn't bring home a copy of the MBX HAL code. Any idea >> when the MBX code will make it into public CVS? > > No need, I have an NT system with the official-release eCos 1.2.3 > available. > > I tried building the MBX stubrom, but the command you suggested > failed as follows, > > $ cygtclsh80.exe ../packages/pkgconf.tcl --target powerpc --platform mbx > --startup rom --disable-kernel --disable uitron --disable-libc > --disable-libm --disable-io --disable-io_serial --disable-wallclock > --disable-watchdog > $ make > $ make -C hal/common/v1_2_3/src/stubrom > //D/PROGRA~1/CYGNUS~1/ECOS/ECOS-M~1/install/lib/libtarget.a(hal_powerpc_hal_in > tr.o): > In function `hal_IRQ_init': > //D/PROGRA~1/CYGNUS~1/ECOS/ECOS-M~1/install/include/cyg/hal/hal_intr.h:525: > undefined reference to `cyg_interrupt_post_dsr' > collect2: ld returned 1 exit status > > Is there a different procedure for 1.2.3? > > I've been following "initialize_stub()" through headers and sources, > and its pretty straight-forward, but I would like to have a > reproducible baseline to start with. > > Thanks again, > -Jamie >