Aren/NetBSD
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
126,761 Commits 606 Branches 0 Tags 3.1 GiB
cdac01064e
Commit Graph

19 Commits

Author SHA1 Message Date
drochner
2209ea5a8a remove license clauses 3 and 4 from my cpoyright notices 2004-03-24 17:11:57 +00:00
bjh21
18a0620f50 Sync acorn26 riscoscalls with acorn32 version. 2003-11-23 15:55:03 +00:00
bjh21
d8b0dd8a97 Create the "machine" and "arm" links directly from the .BEGIN rule, rather 
than in dependencies of it, since .BEGIN is apparently only special for
its commands, not its dependencies.  This should get boot26, and probably
boot32, compiling again.
 2003-11-23 13:34:07 +00:00
lukem
f85d2d1c14 Use ${HOST_SH} instead of `sh'. 
If necessary, pull in <bsd.sys.mk> to get the definition of HOST_SH;
Makefiles that pull in one of (most of) <bsd.*.mk> will get this anyway.
 2003-10-26 07:25:33 +00:00
simonb
970b12080e Define LIB{CRT0,C,CRTBEGIN,CRTEND} as nothing, we don't need to link 
against them.
 2003-10-08 01:40:22 +00:00
matt
cd3cb3dc0f GCC 3.3.1 fixes (-ffreestanding, #include <lib/libkern/libkern.h>) 2003-09-21 06:28:37 +00:00
he
b6643fd4d3 Adapt to const-ification of first arg of open function in fs_ops. 2003-08-21 23:15:18 +00:00
bjh21
610cdc7b39 Create a symlink to sys/arch/arm/include alongside the one to 
sys/arch/${MACHINE}/include.
 2003-04-20 14:37:59 +00:00
bjh21
ae07cdce27 Change the entry pointer of the module from an offset to a branch instruction. 
This allows us to *Load boot26 at 0x8000 and then use *Go to run it.
This might be useful because we can't currently set the RISC OS filetype
for files on CDs, so they all end up being 0xffd (Data), and *RMLoad only
works for files of type 0xffa (Module), so we need some other way of loading
boot26.
 2002-06-13 22:30:22 +00:00
bjh21
b9af29fd4f Increase HIMEM to allow enough space for zlib's tables and 16 MB's worth of 
page tables, so we can boot gzip'ed kernels on 16 MB machines.
 2002-06-13 22:01:44 +00:00
bjh21
40d80d8c16 Add support for LIBSA_NO_FS_CLOSE, LIBSA_NO_FS_WRITE, LIBSA_NO_FS_SEEK and 
LIBSA_NO_TWIDDLE.
LIBSA_NO_FS_SYMLINK is irrelevant here, and LIBSA_FS_SINGLECOMPONENT would
be pointless.
 2002-05-28 14:38:55 +00:00
bjh21
1d76b4db4b 1.4: get_page() abstracted. 2002-05-27 20:18:56 +00:00
bjh21
c489f6804b Add a function, get_page(), whose job is to find which logical page should 
be used to load a given physical page.  At the moment, this just returns
the page's current logical mapping, but in future it might handle returning
bounce pages for physical pages which are in use, and copying to the correct
page just before loading the kernel.
 2002-05-27 20:17:11 +00:00
bjh21
caff9d0e7e Change the various targets to look more like their libsa counterparts. 
I think this'll help on read-only source trees.
 2002-05-05 10:02:17 +00:00
bjh21
505c2b63d1 Re-arrange "machine" magic so it's in the same place as on i386. This seems 
to help things happen in the right order (I smell a missing dependency).
 2002-05-05 10:01:26 +00:00
bjh21
228c4dd3c6 Replace the "machine" logic with something stolen from 
i386/stand/Makefile.booters.  This might make builds with a read-only
source tree happier.
 2002-05-04 12:25:00 +00:00
bjh21
ee7cd026e8 This script has no excuse for ever failing: set -e 2002-04-02 13:17:41 +00:00
bjh21
9597ab134f Add a missing closing quotation mark. NetBSD's /bin/sh coped fine without 
it, but bash (rightly) complained.
 2002-04-02 13:14:43 +00:00
bjh21
1de4f40a8a Rename NetBSD/arm26 to NetBSD/acorn26, so that the two Acorn/ARM ports have 
matching names.  This commit might include some private hacks that have been
lurking in my tree a while.  They're all harmless, and this reduces the number
of gratuitous diffs I have to deal with.
 2002-03-24 15:45:29 +00:00