Re: [OpenBIOS] r116 - openbios-devel/config/scripts
by Segher Boessenkool
> -if test "$host" = "sparc64" -o "$host" = "ia64" -o "$host" = "x86_64"
> -o "$host" = "alpha"; then
> +if test "$host" = "sparc64" -o "$host" = "ia64" -o "$host" = "amd64"
> -o "$host" = "alpha"; then
Do all hosts say this now? Shouldn't you keep x86_64
as well (at least for a while)?
Segher
12 years, 10 months
- 3
- 2
r116 - openbios-devel/config/scripts
by svn＠openbios.org
Author: stepan
Date: 2007-03-23 18:00:01 +0100 (Fri, 23 Mar 2007)
New Revision: 116
Modified:
openbios-devel/config/scripts/crosscflags
Log:
small fix for width detection
Modified: openbios-devel/config/scripts/crosscflags
===================================================================
--- openbios-devel/config/scripts/crosscflags 2007-03-23 16:26:44 UTC (rev 115)
+++ openbios-devel/config/scripts/crosscflags 2007-03-23 17:00:01 UTC (rev 116)
@@ -14,7 +14,7 @@
fi
# host long bits test
-if test "$host" = "sparc64" -o "$host" = "ia64" -o "$host" = "x86_64" -o "$host" = "alpha"; then
+if test "$host" = "sparc64" -o "$host" = "ia64" -o "$host" = "amd64" -o "$host" = "alpha"; then
hostlongbits="64"
else
hostlongbits="32"
@@ -27,7 +27,7 @@
fi
# target long bits test
-if test "$target" = "sparc64" -o "$target" = "ia64" -o "$target" = "x86_64" -o "$target" = "alpha"; then
+if test "$target" = "sparc64" -o "$target" = "ia64" -o "$target" = "amd64" -o "$target" = "alpha"; then
targetlongbits="64"
else
targetlongbits="32"
12 years, 10 months
- 1
- 0
r115 - /
by svn＠openbios.org
Author: stepan
Date: 2007-03-23 17:26:44 +0100 (Fri, 23 Mar 2007)
New Revision: 115
Removed:
openbios-website/
Log:
drop svn based web site, it has been replaced by a wiki since a long time now.
12 years, 10 months
- 1
- 0
Openbios crashes 2
by Jason Hooper
Hi all,
Obviously my last post was not clear enough. I have extracted and compiled
the tar of the previous version in SVN, which is reported as directory
revision 87. I can successfully boot the Debian install iso and the Solaris
install iso starts booting but crashes later, as expected. However, if I
extract and build the latest version, reported as directory revision 114, it
crashes straight away. The exact error reported is "Floating point
exception". This is all on linux 2.6.9.
If I remove the latest patches to obio.c, and then try and boot, I then get
"Unsupported image format".
If I remove the latest patches to esp.c, everything seems to be as it used
to be.
Hope this is of some help.
Regards,
_________________________________________________________________
Solve the Conspiracy and win fantastic prizes.
http://www.theconspiracygame.co.uk/
12 years, 10 months
- 2
- 1
Latest OpenBios crashes...
by Jason Hooper
Hi all,
I have downloaded and built the latest tar.gz file from the repository. It
crashes trying to boot either a Solaris or Linux cdrom with a floating point
exception. Is anybody else getting this or is it only me?
Regards,
_________________________________________________________________
MSN Hotmail is evolving - check out the new Windows Live Mail
http://ideas.live.co.uk
12 years, 10 months
- 2
- 1
Sun OBP bugs in 1.0RC1
by Peter
I've been trying to get old versions of SunOS to load under qemu. In
doing so, I've encountered a number of bugs in OBP. I'm not always
certain of the best fix, but I can at least provide a quick hack that
will get people farther along.
1) Error message: "kmem_alloc failed, nbytes 680"
Bug: obp_dumb_memalloc is a bit too dumb. It needs to pick an address
if passed a null address. (According to the comment in the allocator
in OpenSolaris prom_alloc.c (see
<http://src.opensolaris.org/source/xref/onnv/onnv-gate/usr/src/psm/promif/...>),
"If virthint is zero, a suitable virt is chosen.")
Quick fix: If passed a null address, start doling out addresses at
10MB and increment by size.
Shortcomings: The quick fix ignores the issue of free() and doesn't
remove memory from the virtual-memory/available node.
After the quick fix, the boot gets farther, leading us to:
2) Error message: "Unhandled Exception 0x00000080"
Bug: Trap 0 (entry 0x80 in the table, i.e. syscall_trap_4x) is
undefined. This is because the SunOS bootloader installs the trap by
writing code in the trap table, but the trap table is in the .text
section of OpenBIOS. Thus the trap 0 handler simply jumps to "bug".
Quick fix: Move the trap table to the .data section. Insert a "b
entry; nop; nop; nop;" before "bug:".
Shortcomings: Requires the extra "b entry" code. Allows the only VM
copy of the trap table to be permanently changed. OpenBIOS should
copy the read-only trap table to read-write memory (and update %tbr)
upon reset/entry.
3) #2 above actually exposes another bug. The write to the read-only
trap table does not cause an access violation -- instead, it silently
fails. The "std" instruction at 0x403e6c in the bootloader has no
effect.
Bug: Uncertain. It could be a systemic bug in qemu, but it appears
that the VM's MMU believes that the page is writable. That means that
the VM's MMU is not having the access protection flags set for pages
mapped to ROM. It thinks everything is rwx.
Fix?: The VM's MMU should have the access protection flags properly
set for each ROM section. This should probably be done within
OpenBIOS. E.g., .text should be r-x, .data should probably be rwx,
etc.
This is the one fix I'm really not sure how to implement. Any
suggestions? This may be a problem that only affects this bootloader,
so fixing #2 above may be all that's strictly necessary. But I'm not
positive that this bug doesn't have other ill effects I haven't found
yet.
At any rate, fixing #2 gets us still further, to:
4) Error messages:
"obp_devopen(sd(0,0,0):d) = 0xffd8e270
obp_inst2pkg(fd 0xffd8e270) = 0xffd57f44
obp_getprop(0xffd57f44, device_type) (not found)"
Bug: The OpenBIOS "interpose" implementation is not transparent to
non-interposition-aware code (in violation of the interposition spec).
The inst2pkg call in this sequence returns the phandle for
/packages/misc-files, instead of the proper phandle.
Quick fix: Comment out the "interpose disk-label" lines in ob_sd_open.
Shortcomings: It disables disk-label. The correct fix is to fix the
underlying problem with interposition, but I'm not sure exactly what
it is. Could someone help?
Fixing #4 gets us quite a bit further, until:
5) Error message:
"Unhandled Exception 0x00000009
PC = 0xf0138b20 NPC = 0xf0138b24
Stopping execution"
Bug: The instruction is trying to read from 0xfd020000+4, which is an
invalid address. This address isn't mapped by OBP by default on Sun
hardware, so the bootloader must be trying to (a) map this address and
failing silently or (b) skipping the mapping for some reason. The
instruction is hard-coded to look at this absolute address.
Fix: Unknown. This may be another instance of writes silently
failing, hence my interest in #3 above. It could also be a
side-effect of the quick fix for #4.
I'm happy to work further on these fixes and put them into patch form.
Could someone point me to how I'd do that?
12 years, 10 months
- 4
- 10
x86 ISA binding for open firmware
by (अमेय) Ameya ( पाळंदे) Palande
Hi,
I am not able to locate x86 open firmware isa specific supplements.
Can any one give me information about this ?
I want information about client execution environment on x86.
Thank you.
Regards,
Ameya.
12 years, 10 months
- 3
- 3
problem in running ofclient
by (अमेय) Ameya ( पाळंदे) Palande
Hi,
I am trying to run ofclient on x86.
I started openbios from GRUB.
But its not working, I can't see Hello World message. :(
Can any one tell me how openbios passes client_interface_handler to
client program ?
Thank you,
Regards,
Ameya.
12 years, 10 months
- 2
- 1
Re: [OpenBIOS] Sun OBP bugs in 1.0RC1
by Peter
On 3/8/07, Stefan Reinauer <stepan(a)coresystems.de> wrote:
> So what's the correct thing to do, inst2pkg is supposed to return the
> phandle to sd, even though the device was interposed?
That's correct. The interposition spec
<http://www.openbios.org/data/docs/rec.intpos.d02.pdf> says that
interpose should be "invisible" to interpose-unaware (e.g., IEEE1275
or existing) clients. The SunOS boot loader doesn't know about
interpose, and so expects ihandle>phandle to return the phandle to sd.
It looks like the specified way to query for interposed packages is
via "instance-to-interposed-path", whereas the previously specified
"instance-to-path" omits any reference to interposed packages for
backwards compatibility.
12 years, 10 months
- 1
- 0
r114 - in openbios-devel: arch/sparc32 drivers include/sparc32
by svn＠openbios.org
Author: stepan
Date: 2007-03-09 01:59:05 +0100 (Fri, 09 Mar 2007)
New Revision: 114
Modified:
openbios-devel/arch/sparc32/entry.S
openbios-devel/arch/sparc32/romvec.c
openbios-devel/drivers/esp.c
openbios-devel/drivers/iommu.c
openbios-devel/drivers/obio.c
openbios-devel/include/sparc32/io.h
Log:
Patch for SunOS compatibility from pjcreath+openbios(a)gmail.com:
I've been trying to get old versions of SunOS to load under qemu. In
doing so, I've encountered a number of bugs in OBP. I'm not always
certain of the best fix, but I can at least provide a quick hack that
will get people farther along.
1) Error message: "kmem_alloc failed, nbytes 680"
Bug: obp_dumb_memalloc is a bit too dumb. It needs to pick an address
if passed a null address. (According to the comment in the allocator
in OpenSolaris prom_alloc.c (see
<http://src.opensolaris.org/source/xref/onnv/onnv-gate/usr/src/psm/promif/...>),
"If virthint is zero, a suitable virt is chosen.")
Quick fix: If passed a null address, start doling out addresses at
10MB and increment by size.
Shortcomings: The quick fix ignores the issue of free() and doesn't
remove memory from the virtual-memory/available node.
After the quick fix, the boot gets farther, leading us to:
2) Error message: "Unhandled Exception 0x00000080"
Bug: Trap 0 (entry 0x80 in the table, i.e. syscall_trap_4x) is
undefined. This is because the SunOS bootloader installs the trap by
writing code in the trap table, but the trap table is in the .text
section of OpenBIOS. Thus the trap 0 handler simply jumps to "bug".
Quick fix: Move the trap table to the .data section. Insert a "b
entry; nop; nop; nop;" before "bug:".
Shortcomings: Requires the extra "b entry" code. Allows the only VM
copy of the trap table to be permanently changed. OpenBIOS should
copy the read-only trap table to read-write memory (and update %tbr)
upon reset/entry.
3) #2 above actually exposes another bug. The write to the read-only
trap table does not cause an access violation -- instead, it silently
fails. The "std" instruction at 0x403e6c in the bootloader has no
effect.
Bug: Uncertain. It could be a systemic bug in qemu, but it appears
that the VM's MMU believes that the page is writable. That means that
the VM's MMU is not having the access protection flags set for pages
mapped to ROM. It thinks everything is rwx.
Fix?: The VM's MMU should have the access protection flags properly
set for each ROM section. This should probably be done within
OpenBIOS. E.g., .text should be r-x, .data should probably be rwx,
etc.
This is the one fix I'm really not sure how to implement. Any
suggestions? This may be a problem that only affects this bootloader,
so fixing #2 above may be all that's strictly necessary. But I'm not
positive that this bug doesn't have other ill effects I haven't found
yet.
At any rate, fixing #2 gets us still further, to:
4) Error messages:
"obp_devopen(sd(0,0,0):d) = 0xffd8e270
obp_inst2pkg(fd 0xffd8e270) = 0xffd57f44
obp_getprop(0xffd57f44, device_type) (not found)"
Bug: The OpenBIOS "interpose" implementation is not transparent to
non-interposition-aware code (in violation of the interposition spec).
The inst2pkg call in this sequence returns the phandle for
/packages/misc-files, instead of the proper phandle.
Quick fix: Comment out the "interpose disk-label" lines in ob_sd_open.
Shortcomings: It disables disk-label. The correct fix is to fix the
underlying problem with interposition, but I'm not sure exactly what
it is. Could someone help?
Fixing #4 gets us quite a bit further, until:
5) Error message:
"Unhandled Exception 0x00000009
PC = 0xf0138b20 NPC = 0xf0138b24
Stopping execution"
Bug: The instruction is trying to read from 0xfd020000+4, which is an
invalid address. This address isn't mapped by OBP by default on Sun
hardware, so the bootloader must be trying to (a) map this address and
failing silently or (b) skipping the mapping for some reason. The
instruction is hard-coded to look at this absolute address.
Fix: Unknown. This may be another instance of writes silently
failing, hence my interest in #3 above. It could also be a
side-effect of the quick fix for #4.
6) Error message:
"BAD TRAP: cpu=0 type=9 rp=fd008f0c addr=feff8008 mmu_fsr=3a6 rw=2
MMU sfsr=3a6: Invalid Address on supv data store at level 3
regs at fd008f0c:
psr=4400fc7 pc=f00053f4 npc=f00053f8
..."
Bug: Real sun4m hardware registers 4 CPU-specific interrupts followed
by a system-wide interrupt, regardless of the number of CPUs
installed. The same is true of counters. SunOS looks at the 5th
interrupt for the system-wide interrupt. OBP, since there's only one
CPU, just sets up one CPU-specific interrupt followed by the
system-wide interrupt, so there is no 5th interrupt. See the comment
on "NCPU" at
<http://stuff.mit.edu/afs/athena/astaff/project/opssrc/sys.sunos/sun4m/dev...>.
Fix: in obp_interrupt_init() and obp_counter_init() register 4
CPU-specific interrupts before allocating the system-wide interrupt.
The kernel will then map the 5th interrupt to the system-wide
interrupt.
7) Error message:
"BAD TRAP: cpu=0 type=9 rp=fd008d8c addr=7ff000 mmu_fsr=126 rw=1
MMU sfsr=126: Invalid Address on supv data fetch at level 1
regs at fd008d8c:
psr=4000cc4 pc=f01339a4 npc=f01339a8
..."
Bug: The command-line arguments passed to the kernel are fixed at
address 0x7FF000 (CMDLINE_ADDR, passed from qemu via nv_info.cmdline),
which is no longer mapped by the time the kernel looks at the boot
arguments. A regular Sun boot ROM will copy this into mapped memory.
Fix: Copy the string in nv_info.cmdline to a OpenBIOS global (since
OpenBIOS continues to be mapped) in ob_nvram_init().
8) Error message:
"BAD TRAP: cpu=0 type=9 rp=fd008dec addr=1019000 mmu_fsr=126 rw=1
MMU sfsr=126: Invalid Address on supv data fetch at level 1
regs at fd008dec:
psr=4400cc5 pc=f0131680 npc=f0131684
..."
Bug: The dumb memory allocator from bug #1 was allocating a range that
the SunOS 4 kernel doesn't like.
Fix: Mimic the Sun boot ROM allocator: the top of the heap should be
a 0xFFEDA000 and allocations should return descending addresses. So,
for example, if asking for 0x1000 bytes, the first returned pointer
should be 0xFFED9000.
9) Error message:
"BAD TRAP: cpu=0 type=9 rp=fd008d2c addr=b1b91000 mmu_fsr=126 rw=1
MMU sfsr=126: Invalid Address on supv data fetch at level 1
regs at fd008d2c:
psr=4900cc3 pc=f0142c04 npc=f0142c08
..."
Bug: The precise underlying cause isn't clear. The bug appears due to
a variation between OBP's behavior and stock Sun behavior.
Fix: Add the "cache-physical?" property to the CPU node in
ob_nvram_init() and bump the "mmu-nctx" property up to 4096 (from
256).
Modified: openbios-devel/arch/sparc32/entry.S
===================================================================
--- openbios-devel/arch/sparc32/entry.S 2006-10-30 23:06:43 UTC (rev 113)
+++ openbios-devel/arch/sparc32/entry.S 2007-03-09 00:59:05 UTC (rev 114)
@@ -60,15 +60,15 @@
sta %g1, [%g0] ASI_M_MMUREGS ! enable mmu
first_cpu:
+ /* Create temporary page tables and map the ROM area to end of
+ RAM. This will be done properly in iommu.c later. */
set PHYS_JJ_EEPROM + 0x30, %g1
lda [%g1] ASI_M_BYPASS, %g1
- ! map PROLDATA to PROLBASE+PROLSIZE to end of ram
- !set PROLSIZE+0x1000-PROLDATA+PROLBASE, %g2 ! add 0x1000 for temp tables
set _end, %g3
set 0xfff, %g2
add %g3, %g2, %g3
andn %g3, %g2, %g3
- set _data, %g2
+ set _start, %g2
sub %g3, %g2, %g3
set 0x1000, %g4 ! add 0x1000 for page tables
add %g4, %g3, %g2
@@ -128,45 +128,17 @@
sta %g3, [%g2] ASI_M_BYPASS
add %g2, 0xa00-0x9e0, %g2 ! s+0xa00
- !set (PROLDATA-PROLBASE)/0x1000, %g6 ! # of .text pages
- set _start, %g3
- set _data, %g6
- set 0x1000, %g5
- sub %g6, %g3, %g6
- srl %g6, 12, %g6 ! # of .text pages
-1: srl %g3, 0x4, %g4
- or %g4, ((7 << 2) | 2), %g4 ! 4 = U: --X S: --X (rom, execute only)
- sta %g4, [%g2] ASI_M_BYPASS
- add %g2, 4, %g2
- add %g3, %g5, %g3
- deccc %g6
- bne 1b
- nop
-#if 0
- !set (PROLDATA-PROLRODATA)/0x1000, %g6 ! # of .rodata pages
- set _data, %g6
- set _rodata, %g4
- sub %g6, %g4, %g6
- srl %g6, 12, %g6 ! # of .bss pages
-1: srl %g3, 0x4, %g4
- or %g4, ((0 << 2) | 2), %g4 ! 0 = U: R-- S: R-- (rom, read only)
- sta %g4, [%g2] ASI_M_BYPASS
- add %g2, 4, %g2
- add %g3, %g5, %g3
- deccc %g6
- bne 1b
- nop
-#endif
- !set (PROLBASE+PROLSIZE-PROLDATA)/0x1000, %g6 ! # of .bss pages
+ /* Use end of ram for code, rodata, data, and bss
+ sections. SunOS wants to write to trap table... */
set _end, %g6
- set _data, %g4
+ set _start, %g4
sub %g6, %g4, %g6
- srl %g6, 12, %g6 ! # of .bss pages
- set 0x1000, %g4
+ srl %g6, 12, %g6 ! # of all pages
+ set 0x1000, %g5
sll %g7, 0x4, %g3
- add %g4, %g3, %g3
+ add %g5, %g3, %g3 ! ctx table + 0x1000
1: srl %g3, 0x4, %g4
- or %g4, ((7 << 2) | 2), %g4 ! 5 = U: R-- S: RW- (data area, read/write)
+ or %g4, ((7 << 2) | 2), %g4 ! 7 = U: --- S: RWX
sta %g4, [%g2] ASI_M_BYPASS
add %g2, 4, %g2
add %g3, %g5, %g3
@@ -174,8 +146,26 @@
bne 1b
nop
- mov %g1, %g3
+ mov %g1, %g6 ! %g6 = memory size
+ /* Copy the code, rodata and data sections from ROM. */
+ set 0x1000 - 4, %g4
+ sll %g7, 0x4, %g3
+ add %g4, %g3, %g3 ! ctx table + 0x1000 - 4
+ set _start - 4, %g4 ! First address of TEXT
+ set _bss, %g5 ! Last address of DATA
+ ba 2f
+ nop
+1:
+ lda [%g4] ASI_M_BYPASS, %g1
+ sta %g1, [%g3] ASI_M_BYPASS
+2:
+ cmp %g4, %g5
+ add %g3, 0x4, %g3
+ bl 1b
+ add %g4, 0x4, %g4
+
+
set AC_M_CTPR, %g2
sta %g7, [%g2] ASI_M_MMUREGS ! set ctx table ptr
set 1, %g1
@@ -196,19 +186,6 @@
wr %g2, 0x0, %psr
WRITE_PAUSE
- /* Copy the DATA section from ROM. */
- set _data - 4, %o0 ! First address of DATA
- set _bss, %o1 ! Last address of DATA
- ba 2f
- nop
-1:
- lda [%o0] ASI_M_BYPASS, %g1
- st %g1, [%o0]
-2:
- subcc %o0, %o1, %g0
- bl 1b
- add %o0, 0x4, %o0
-
/* Zero out our BSS section. */
set _bss - 4, %o0 ! First address of BSS
set _estack, %o1 ! Last address of BSS
@@ -229,10 +206,10 @@
wr %g1, 0x0, %tbr
set qemu_mem_size, %g1
- st %g3, [%g1]
+ st %g6, [%g1]
sll %g7, 4, %g7 ! Store va->pa conversion factor
- set _data - 0x1000, %g1
+ set _start - 0x1000, %g1
sub %g1, %g7, %g7
set va_shift, %g1
st %g7, [%g1]
Modified: openbios-devel/arch/sparc32/romvec.c
===================================================================
--- openbios-devel/arch/sparc32/romvec.c 2006-10-30 23:06:43 UTC (rev 113)
+++ openbios-devel/arch/sparc32/romvec.c 2007-03-09 00:59:05 UTC (rev 114)
@@ -108,6 +108,21 @@
}
}
+#if CONFIG_DEBUG_OBP
+static int looks_like_string(char *str, int len)
+{
+ int i;
+ int ret = (str[len-1] == '\0');
+ for (i = 0; i < len-1 && ret; i++)
+ {
+ int ch = str[i] & 0xFF;
+ if (ch < 0x20 || ch > 0x7F)
+ ret = 0;
+ }
+ return ret;
+}
+#endif
+
static int obp_getprop(int node, char *name, char *value)
{
int notfound, found;
@@ -148,7 +163,19 @@
else
str = "NULL";
- DPRINTF("obp_getprop(0x%x, %s) = %s\n", node, name, str);
+#if CONFIG_DEBUG_OBP
+ if (looks_like_string(str, len)) {
+ DPRINTF("obp_getprop(0x%x, %s) = %s\n", node, name, str);
+ } else {
+ int i;
+ DPRINTF("obp_getprop(0x%x, %s) = ", node, name);
+ for (i = 0; i < len; i++) {
+ DPRINTF("%02x%s", str[i] & 0xFF,
+ (len == 4 || i == len-1) ? "" : " ");
+ }
+ DPRINTF("\n");
+ }
+#endif
return len;
}
@@ -378,8 +405,22 @@
static char * obp_dumb_memalloc(char *va, unsigned int size)
{
+ static unsigned int next_free_address = 0xFFEDA000;
+
free_ram -= size;
DPRINTF("obp_dumb_memalloc req 0x%x of %d at 0x%x\n", va, size, free_ram);
+
+ // If va is null, the allocator is supposed to pick a "suitable" address.
+ // (See OpenSolaric prom_alloc.c) There's not any real guidance as
+ // to what might be "suitable". So we mimic the behavior of a Sun boot
+ // ROM.
+
+ if (va == NULL) {
+ va = next_free_address - size;
+ next_free_address -= size;
+ DPRINTF("obp_dumb_memalloc req null -> 0x%x\n", va);
+ }
+
obp_dumb_mmap(va, 1, free_ram, size);
return va;
@@ -447,7 +488,7 @@
void *
init_openprom(unsigned long memsize)
{
- free_ram = va2pa((int)&_data) - PAGE_SIZE;
+ free_ram = va2pa((int)&_start) - PAGE_SIZE;
ptphys = totphys;
ptmap = totmap;
Modified: openbios-devel/drivers/esp.c
===================================================================
--- openbios-devel/drivers/esp.c 2006-10-30 23:06:43 UTC (rev 113)
+++ openbios-devel/drivers/esp.c 2007-03-09 00:59:05 UTC (rev 114)
@@ -285,12 +285,24 @@
selfword("open-deblocker");
+#if WHEN_INTERPOSE_IS_FIXED
+ // Buggy implementation of "interpose" breaks SunOS 4.
+ // inst2pkg of the device returned by open() should return
+ // the same phandle regardless of whether any interposition
+ // has happened. Interpose currently causes inst2pkg to
+ // return the phandle of /packages/misc-files.
+ //
+ // Removing interpose isn't the ideal solution, but it
+ // allows SunOS 4 to load the kernel. Interpose should
+ // be fixed, then re-enabled.
+
/* interpose disk-label */
ph = find_dev("/packages/disk-label");
fword("my-args");
PUSH_ph( ph );
fword("interpose");
-
+#endif
+
RET ( -ret );
}
Modified: openbios-devel/drivers/iommu.c
===================================================================
--- openbios-devel/drivers/iommu.c 2006-10-30 23:06:43 UTC (rev 113)
+++ openbios-devel/drivers/iommu.c 2007-03-09 00:59:05 UTC (rev 114)
@@ -222,7 +222,7 @@
PUSH(0);
fword("encode-int");
fword("encode+");
- PUSH(va2pa((unsigned long)&_data) - PAGE_SIZE);
+ PUSH(va2pa((unsigned long)&_start) - PAGE_SIZE);
fword("encode-int");
fword("encode+");
push_str("available");
@@ -295,14 +295,8 @@
l1[i] = SRMMU_ET_INVALID;
}
- // 1:1 mapping for ROM
- pa = va = (unsigned long)&_start;
- for (; va < (unsigned long)&_data; va += PAGE_SIZE, pa += PAGE_SIZE) {
- map_page(va, pa, 0);
- }
-
- // data & bss mapped to end of RAM
- va = (unsigned long)&_data;
+ // text, rodata, data, and bss mapped to end of RAM
+ va = (unsigned long)&_start;
for (; va < (unsigned long)&_end; va += PAGE_SIZE) {
pa = va2pa(va);
map_page(va, pa, 0);
Modified: openbios-devel/drivers/obio.c
===================================================================
--- openbios-devel/drivers/obio.c 2006-10-30 23:06:43 UTC (rev 113)
+++ openbios-devel/drivers/obio.c 2007-03-09 00:59:05 UTC (rev 114)
@@ -37,6 +37,20 @@
#define PROMDEV_SCREEN 0 /* output to screen */
#define PROMDEV_TTYA 1 /* in/out to ttya */
+/* "NCPU" is an historical name that's now a bit of a misnomer. The sun4m
+ * architecture registers NCPU CPU-specific interrupts along with one
+ * system-wide interrupt, regardless of the number of actual CPUs installed.
+ * See the comment on "NCPU" at <http://stuff.mit.edu/afs/athena/astaff/
+ * project/opssrc/sys.sunos/sun4m/devaddr.h>.
+ */
+#define SUN4M_NCPU 4
+
+/* The kernel may want to examine the arguments, so hold a copy in OBP's
+ * mapped memory.
+ */
+#define OBIO_CMDLINE_MAX 256
+char obio_cmdline[OBIO_CMDLINE_MAX];
+
/* DECLARE data structures for the nodes. */
DECLARE_UNNAMED_NODE( ob_obio, INSTALL_OPEN, sizeof(int) );
@@ -306,6 +320,7 @@
const char *stdin, *stdout;
unsigned int i;
char nographic;
+ uint32_t size;
ob_new_obio_device("eeprom", NULL);
@@ -325,8 +340,14 @@
}
kernel_image = nv_info.kernel_image;
kernel_size = nv_info.kernel_size;
- cmdline = nv_info.cmdline;
- cmdline_size = nv_info.cmdline_size;
+
+ size = cmdline_size;
+ if (size > OBIO_CMDLINE_MAX) size = OBIO_CMDLINE_MAX;
+ memcpy(obio_cmdline, nv_info.cmdline, size);
+ obio_cmdline[size-1] = '\0';
+ cmdline = obio_cmdline;
+ cmdline_size = size;
+
boot_device = nv_info.boot_device;
nographic = nv_info.nographic;
@@ -369,6 +390,10 @@
push_str("version");
fword("property");
+ push_str("");
+ push_str("cache-physical?");
+ fword("property");
+
PUSH(32);
fword("encode-int");
push_str("cache-line-size");
@@ -435,7 +460,7 @@
push_str("ncaches");
fword("property");
- PUSH(256);
+ PUSH(4096);
fword("encode-int");
push_str("mmu-nctx");
fword("property");
@@ -539,26 +564,32 @@
ob_counter_init(unsigned long base, unsigned long offset)
{
volatile struct sun4m_timer_regs *regs;
+ int i;
ob_new_obio_device("counter", NULL);
+ for (i = 0; i < SUN4M_NCPU; i++) {
+ PUSH(0);
+ fword("encode-int");
+ if (i != 0) fword("encode+");
+ PUSH(offset + (i * PAGE_SIZE));
+ fword("encode-int");
+ fword("encode+");
+ PUSH(COUNTER_REGS);
+ fword("encode-int");
+ fword("encode+");
+ }
+
PUSH(0);
fword("encode-int");
- PUSH(offset);
- fword("encode-int");
fword("encode+");
- PUSH(COUNTER_REGS);
- fword("encode-int");
- fword("encode+");
- PUSH(0);
- fword("encode-int");
- fword("encode+");
PUSH(offset + 0x10000);
fword("encode-int");
fword("encode+");
PUSH(COUNTER_REGS);
fword("encode-int");
fword("encode+");
+
push_str("reg");
fword("property");
@@ -580,26 +611,32 @@
static void
ob_interrupt_init(unsigned long base, unsigned long offset)
{
+ int i;
ob_new_obio_device("interrupt", NULL);
+ for (i = 0; i < SUN4M_NCPU; i++) {
+ PUSH(0);
+ fword("encode-int");
+ if (i != 0) fword("encode+");
+ PUSH(offset + (i * PAGE_SIZE));
+ fword("encode-int");
+ fword("encode+");
+ PUSH(INTERRUPT_REGS);
+ fword("encode-int");
+ fword("encode+");
+ }
+
PUSH(0);
fword("encode-int");
- PUSH(offset);
- fword("encode-int");
fword("encode+");
- PUSH(INTERRUPT_REGS);
- fword("encode-int");
- fword("encode+");
- PUSH(0);
- fword("encode-int");
- fword("encode+");
PUSH(offset + 0x10000);
fword("encode-int");
fword("encode+");
PUSH(INTERRUPT_REGS);
fword("encode-int");
fword("encode+");
+
push_str("reg");
fword("property");
Modified: openbios-devel/include/sparc32/io.h
===================================================================
--- openbios-devel/include/sparc32/io.h 2006-10-30 23:06:43 UTC (rev 113)
+++ openbios-devel/include/sparc32/io.h 2007-03-09 00:59:05 UTC (rev 114)
@@ -10,7 +10,7 @@
static inline unsigned long
va2pa(unsigned long va)
{
- if ((va >= (unsigned long)&_data) &&
+ if ((va >= (unsigned long)&_start) &&
(va < (unsigned long)&_end))
return va - va_shift;
else
@@ -20,8 +20,8 @@
static inline unsigned long
pa2va(unsigned long pa)
{
- if ((pa + va_shift >= (unsigned long)&_data) &&
- (pa + va_shift< (unsigned long)&_end))
+ if ((pa + va_shift >= (unsigned long)&_start) &&
+ (pa + va_shift < (unsigned long)&_end))
return pa + va_shift;
else
return pa;
12 years, 10 months
- 1
- 0