basilisk-devel — Basilisk II development-related discussion

[B2-devel] Archive List

[Mark H. <ma...@mm...>]

Hello all,

Sorry for this email but my ISP filled bankruptcy and put my digital 
life in total caos for the last few days.  The last list email I got was 
around the 24th.  Is there an archive that I can go back and read what 
I've missed so far?

Thanks!
Mark

[B2-devel] Sound support with NewWorld ROMs?

[Gwenole B. <gb...@di...>]

Hi,

I still fail to see why audio output doesn't work with G3+ based ROMs 
(including NewWorld). There is no audio sifter built in the ROM. 
However, MacOS 8.6 loads 'thng' based sifters. Recently, I patched the 
GetResource() thunk to correctly sign-extend the resource ID. This 
makes check_load_invoc() recognize and add those sifters. However, they 
don't seem to be enabled and things work as if we are in "mute" mode, 
though the sound level is at its maximum.

Christian, would you know how to fix that? You are the in-depth MacOS 
expert. ;-)

Probably colateral note: I noticed that MOL makes MacOS 8.6 load an 
extra enabler (MacOS ROM enabler or something like that). SheepShaver 
doesn't, thus probably making MacOS to complain that the DisplayEnabler 
is not loaded when one tries to open the "Monitors" control panel. 
Maybe that "macroscopic" enabler includes the DisplayEnabler and 
something like an AudioEnabler?

Bye,
Gwenole.

Re: [B2-devel] Improved SheepShaver/JIT

[Robert S. <Ro...@xb...>]

I have noticed an approx 13% boost in speed for real world bench marks.
Unstuffing files , loading large images in Graphics converter and
preforming fliters on those images. By using ICC over GCC using the -O3
arugement.  This was on a P4 Northwood.  I will have to see what kind of
boost i get now with your updated code when my p4 machine is among the
living .  My P4 died my Athlon died my G4s 250 gig drive die all in the
weekend.  I will be able to get the athlon machine up first tas all it
needs is a new bios chip.  But I am going to download sheep shaver from
cvs again and try it out on this old machine just to see the speed it
pulling off of an old 600mhz celey.



On Tue, 2004-01-27 at 09:03, Gwenole Beauchesne wrote:
> Hi,
> 
> I have just committed a few changes to the SheepShaver JIT. Basically,
> gprof told me that 64% of the time was spent in block_cache::find().  
> Besides, some extra counters revealed that more than 95% of the blocks are
> available at the head of the blocks list for a hash table entry.
> 
> i.e. I inlined that fast block lookup case in generated code. On a
> Pentium4, Speedometer 4 pure "CPU" benchtest improved by a factor of 3.  
> More testing/tuning has yet to be done but overall the speed increase is
> interesting. ;-)
> 
> Please give it a try.
> 
> BTW, there are also fixes to enable MacBench 5.0 to work.
> 
> Bye,
> Gwenole.
> 
> 
> -------------------------------------------------------
> The SF.Net email is sponsored by EclipseCon 2004
> Premiere Conference on Open Tools Development and Integration
> See the breadth of Eclipse activity. February 3-5 in Anaheim, CA.
> http://www.eclipsecon.org/osdn
> _______________________________________________
> basilisk-devel mailing list
> bas...@li...
> https://lists.sourceforge.net/lists/listinfo/basilisk-devel

Re: [B2-devel] Improved SheepShaver/JIT

[Gwenole B. <gb...@di...>]

On Tue, 27 Jan 2004, Gwenole Beauchesne wrote:

> BTW, there are also fixes to enable MacBench 5.0 to work.

Some results:		CPU	FPU (relative to a G3/300)
- P4 @ 1.8 GHz: 	10%	25%
- Opteron @ 1.8 GHz:	21%	89%

I will check integer bottleneck later. And yes, AMD64 platforms are always 
more interesting. ;-)

[B2-devel] Improved SheepShaver/JIT

[Gwenole B. <gb...@di...>]

Hi,

I have just committed a few changes to the SheepShaver JIT. Basically,
gprof told me that 64% of the time was spent in block_cache::find().  
Besides, some extra counters revealed that more than 95% of the blocks are
available at the head of the blocks list for a hash table entry.

i.e. I inlined that fast block lookup case in generated code. On a
Pentium4, Speedometer 4 pure "CPU" benchtest improved by a factor of 3.  
More testing/tuning has yet to be done but overall the speed increase is
interesting. ;-)

Please give it a try.

BTW, there are also fixes to enable MacBench 5.0 to work.

Bye,
Gwenole.

[B2-devel] Re: basilisk-devel digest, Vol 1 #257 - 1 msg

[Mark H. <ma...@mm...>]

----- Original Message ----- 
From: <bas...@li...>
To: <bas...@li...>
Sent: Saturday, January 24, 2004 11:01 PM
Subject: basilisk-devel digest, Vol 1 #257 - 1 msg


> Send basilisk-devel mailing list submissions to
> bas...@li...
>
> To subscribe or unsubscribe via the World Wide Web, visit
> https://lists.sourceforge.net/lists/listinfo/basilisk-devel
> or, via email, send a message with subject or body 'help' to
> bas...@li...
>
> You can reach the person managing the list at
> bas...@li...
>
> When replying, please edit your Subject line so it is more specific
> than "Re: Contents of basilisk-devel digest..."
>
>
> Today's Topics:
>
>    1. Re: Zaurus SheepShaver with gcc33 (Gwenole Beauchesne)
>
> --__--__--
>
> Message: 1
> Date: Sat, 24 Jan 2004 12:48:15 +0100
> Subject: Re: [B2-devel] Zaurus SheepShaver with gcc33
> From: Gwenole Beauchesne <gb...@di...>
> To: bas...@li...
> Reply-To: bas...@li...
>
> Hi,
>
> > ../kpx_cpu/src/cpu/ppc/ppc-translate.cpp:99: error: (Each undeclared
> > identifier
> >    is reported only once for each function it appears in.)
> > make: *** [obj/ppc-translate.o] Error 1
>
> Move up the #if PPC_ENABLE_JIT as there is no JIT support for ARM yet,
> nor disassembler.
>
> BTW, where is the gcc33 env on ipaq4?
>


Hi Gwenole,

I thought that that autogen set the no jit compiler option once it saw that
it was compiling for a non-x86 cpu?  I'm pretty sure I remember seeing the
enable jit compiler option set to no in the config summary.  I could be
wrong so I'll double check.

If you go into /home/guest/dso...@ma... and browse you will see
three env scripts to set either gcc 2.9 something, 3.2.1, and then 3.3.
Feel free to use these or anything in in the dir!

Thanks for all the help,
Mark

Re: [B2-devel] Zaurus SheepShaver with gcc33

[Gwenole B. <gb...@di...>]

Hi,

> ../kpx_cpu/src/cpu/ppc/ppc-translate.cpp:99: error: (Each undeclared
> identifier
>    is reported only once for each function it appears in.)
> make: *** [obj/ppc-translate.o] Error 1

Move up the #if PPC_ENABLE_JIT as there is no JIT support for ARM yet, 
nor disassembler.

BTW, where is the gcc33 env on ipaq4?

[B2-devel] Zaurus SheepShaver with gcc33

[Mark H. <ma...@mm...>]

That's a compiler error, you would need a more recent one than 2.95.4
currently installed on ipaq4. Someone appears to have built a gcc 3.0.4
himself, available in ~/suba/InstGcc304/bin.
You could try with it by exporting proper CC & CXX variables. Note:
make distclean first.

Bye,
Gwenole.

Hi Gwenole,

I was running the wrong env file so I configured it to gcc33 and it gets
much farther!  :-)

However, it does bomb out here:

g++3.3 -I../kpx_cpu/include -I../kpx_cpu/src -I../include -I. -DHAVE_CONFIG_
H -D
_REENTRANT -DDATADIR=\"/usr/local/share/SheepShaver\" -g -O2  -I/usr/X11R6/i
nclu
de -I/usr/include/gtk-1.2 -I/usr/include/glib-1.2 -I/usr/lib/glib/include  -
c ..
/kpx_cpu/src/cpu/ppc/ppc-execute.cpp -o obj/ppc-execute.o
g++3.3 -I../kpx_cpu/include -I../kpx_cpu/src -I../include -I. -DHAVE_CONFIG_
H -D
_REENTRANT -DDATADIR=\"/usr/local/share/SheepShaver\" -g -O2  -I/usr/X11R6/i
nclu
de -I/usr/include/gtk-1.2 -I/usr/include/glib-1.2 -I/usr/lib/glib/include  -
c ..
/kpx_cpu/src/cpu/ppc/ppc-translate.cpp -o obj/ppc-translate.o
../kpx_cpu/src/cpu/ppc/ppc-translate.cpp: In function `void
   disasm_translation(unsigned int, unsigned int, uint8*, unsigned int)':
../kpx_cpu/src/cpu/ppc/ppc-translate.cpp:99: error: `TARGET_NATIVE'
undeclared
   (first use this function)
../kpx_cpu/src/cpu/ppc/ppc-translate.cpp:99: error: (Each undeclared
identifier
   is reported only once for each function it appears in.)
make: *** [obj/ppc-translate.o] Error 1

Any thoughts?

Thanks!
Mark

Re: [B2-devel] SheepShaver on Zaurus

[Gwenole B. <gb...@di...>]

Hi,

> ../kpx_cpu/src/cpu/ppc/ppc-decode.cpp:113: Internal compiler error.
> ../kpx_cpu/src/cpu/ppc/ppc-decode.cpp:113: Please submit a full bug 
> report.
> ../kpx_cpu/src/cpu/ppc/ppc-decode.cpp:113: See 
> <URL:http://www.gnu.org/software/
> gcc/bugs.html> for instructions.
> make: *** [obj/ppc-decode.o] Error 1

That's a compiler error, you would need a more recent one than 2.95.4 
currently installed on ipaq4. Someone appears to have built a gcc 3.0.4 
himself, available in ~/suba/InstGcc304/bin.
You could try with it by exporting proper CC & CXX variables. Note: 
make distclean first.

Bye,
Gwenole.

[B2-devel] SheepShaver on Zaurus

[Mark H. <ma...@mm...>]

Update on compiling SheepShaver for the Zaurus:

I get this error:

c++ -I../kpx_cpu/include -I../kpx_cpu/src -I../include -I. -DHAVE_CONFIG_H -
D_RE
ENTRANT -DDATADIR=\"/usr/local/share/SheepShaver\" -g -O2  -I/usr/X11R6/incl
ude
-I/usr/include/gtk-1.2 -I/usr/include/glib-1.2 -I/usr/lib/glib/include  -c
../kp
x_cpu/src/cpu/ppc/ppc-decode.cpp -o obj/ppc-decode.o
../kpx_cpu/src/cpu/ppc/ppc-decode.cpp:113: Internal compiler error.
../kpx_cpu/src/cpu/ppc/ppc-decode.cpp:113: Please submit a full bug report.
../kpx_cpu/src/cpu/ppc/ppc-decode.cpp:113: See
<URL:http://www.gnu.org/software/
gcc/bugs.html> for instructions.
make: *** [obj/ppc-decode.o] Error 1

You can view the entire log here: http://www.mmhart.com/sslog.txt

If anybody has any comments that would be great, if not its no big deal, I
know there are bigger issues out there.  :-)  Heck, I'm just excited that
somebody is working on a PPC emulator!

Thanks,
Mark

[B2-devel] SheepShaver on Zaurus?

[Mark H. <ma...@mm...>]

Hi Gwenole,

SheepShaver may now build and possibly run on Zaurus/ARM PDAs.

Awesome!  I compiled Basilisk II for the Z and it runs great!  I tried =
SS but I had some compile issues.  I'll try it again with your new build =
and let you know how it goes.

I'm not expecting any great speed with the XScale @ 400mhz but I think =
it would be cool to run.  Howeve, B2 actually does run pretty nice and =
the Z makes a great Mac PDA.  :-)

Thanks,
Mark

Re: [B2-devel] Darwin port

[Christian B. <Chr...@un...>]

Hi!

On Thu, Jan 22, 2004 at 07:48:53PM +0900, Toshimitsu Tanaka wrote:
> -	{STR_VOLUME_IS_MOUNTED_WARN, "The volume '%s' is mounted under 
> Linux.  Basilisk II will try to unmount it."},
> +#if defined(__APPLE__) && defined(__MACH__)
> +        {STR_VOLUME_IS_MOUNTED_WARN, "The volume '%s' is mounted under  
> Darwin. SheepShaver will try to unmount it."},
> [...]

This is somewhat ugly and would have to be changed again for *BSD, Solaris,
etc. Maybe it should use the $target_os from the configure script or some
more "user friendly" OS name that is selected in sysdeps.h, or simply "Unix"
as B2 (which is admittedly a little too generic, though).

Bye,
Christian

-- 
  / Physics is an algorithm
\/ http://www.uni-mainz.de/~bauec002/

[B2-devel] SheepShaver Ethernet weirdness

[Gwenole B. <gb...@di...>]

Hi,

I have arranged SheepShaver code to be a little more little-endian aware. 
Current patches are available here for review:
<http://gwenole.beauchesne.free.fr/sheepshaver/patches/>
File: sheep_ethernet.diff

However, with highly loaded pages (live.apple.com or whatever is the
default page for MSIE 5.1), SheepShaver/x86 will hang because it can no
longer allocb() enough space (WARNING: Cannot allocate mblk for received
packet).  Then, I wondered, when are we supposed to call freeb()? I
looked, we never do so. I think freemsg() will call it though.

I haven't tried on native/ppc yet, so I was wondering whether my patches
didn't fully make little-endian clean code, or is there another problem in
SheepShaver core. I am using sheep_net ethernet support.

Bye,
Gwenole.

Re: [B2-devel] Darwin port

[Toshimitsu T. <t_t...@db...>]

Hi.

On 2004/01/19, at 8:26, Gwenole Beauchesne wrote:
> Darwin bits are now committed. Tested on MacOS X 10.2 with Apple X11  
> server. Note you may notice that SheepShaver doesn't fully exit on  
> shutdown for unknown reason at the moment.
>
>> The Darwin port is almost working (to be committed):
>> <http://gwenole.beauchesne.free.fr/sheepshaver/ 
>> capture_darwin_port2.pdf>

Good job.
I tried it on 10.3.
http://member.nifty.ne.jp/poseidon/images/ss_x.jpg

But, ExtFS Volume Name is "Linux"...



--- SheepShaver/src/Unix/user_strings_unix.cpp.org	Thu Jan 22 18:45:00  
2004
+++ SheepShaver/src/Unix/user_strings_unix.cpp	Thu Jan 22 19:43:48 2004
@@ -25,10 +25,22 @@
  // Platform-specific string definitions
  user_string_def platform_strings[] = {
  	// Common strings that have a platform-specific variant
-	{STR_VOLUME_IS_MOUNTED_WARN, "The volume '%s' is mounted under Linux.  
Basilisk II will try to unmount it."},
+#if defined(__APPLE__) && defined(__MACH__)
+        {STR_VOLUME_IS_MOUNTED_WARN, "The volume '%s' is mounted under  
Darwin. SheepShaver will try to unmount it."},
+        {STR_EXTFS_CTRL, "Darwin Root"},
+        {STR_EXTFS_NAME, "Darwin Directory Tree"},
+        {STR_EXTFS_VOLUME_NAME, "Darwin"},
+#elif defined(__linux__)
+        {STR_VOLUME_IS_MOUNTED_WARN, "The volume '%s' is mounted under  
Linux. SheepShaver will try to unmount it."},
  	{STR_EXTFS_CTRL, "Linux Root"},
  	{STR_EXTFS_NAME, "Linux Directory Tree"},
-	{STR_EXTFS_VOLUME_NAME, "Linux"},
+        {STR_EXTFS_VOLUME_NAME, "Linux"},
+#else
+        {STR_VOLUME_IS_MOUNTED_WARN, "The volume '%s' is mounted under  
Unix. SheepShaver will try to unmount it."},
+        {STR_EXTFS_CTRL, "Unix Root"},
+        {STR_EXTFS_NAME, "Unix Directory Tree"},
+	{STR_EXTFS_VOLUME_NAME, "Unix"},
+#endif

  	// Purely platform-specific strings
  	{STR_NO_DEV_ZERO_ERR, "Cannot open /dev/zero: %s."},
---
Toshimitsu Tanaka
t_t...@db...
http://member.nifty.ne.jp/poseidon/index.html

My SheepShaver page:
http://member.nifty.ne.jp/poseidon/emu/sheepshaver1.html
(Sorry,Japanese text only)

[B2-devel] SheepShaver on Zaurus?

[Gwenole B. <gb...@di...>]

Hi,

SheepShaver may now build and possibly run on Zaurus/ARM PDAs. I don't 
have such beast but I only added necessary support to sigsegv.cpp to 
skip instructions causing a fault. If there were any ARM expert that 
could proofread the load/store emulation, that would be fine. Though, 
usual configure tests pass when even compiled with -march=armv4 
(halfword load/store support).

BTW, how fast is an Xscale @ 400 MHz?

Bye,
Gwenole.

Re: [B2-devel] Darwin port

[Michael S. <msl...@co...>]

Gwenole,

> Actually, the sigsegv handler code is the same as for B2. I have just 
> hacked around to use Mach exceptions to handle screen faults (early so 
> that we don't need the whole context) + other faults like writes to 
> ROM. However, SheepShaver hangs early after a few seconds. I think 
> this is because we have not taught our Mach exception handler to 
> execute with an alternate stack (set by sigaltstack() or another means 
> we can define) as we should. Is that possible to do?


The short answer is no it is not possible in any way I know.

I really would not be using sigaltstack() if I were you. It dates from 
old BSDs. The reason it ever existed is that some signal handlers in 
other systems got not only the signal number but two other args, one of 
which was a pointer into a structure that contained cpu context. In 
these old versions of BSD you had no easy way of finding where your 
context was on the stack, later versions had header files that helped. 
There are also times when you really do need an alternate stack for 
your signals from your normal stack, so it is good to use there. But in 
any case I would not use sigaltstack() for multi-threaded programs. 
Also as you may be seeing, the interaction between exceptions and the 
alternate signal stack is probably not something that Apple ever 
debugged.

Now the long answer to why you cannot put the exception handler on the 
alternate signal stack. What happen in Mach exceptions is that you have 
some thread somewhere that handles the exceptions. It can be in another 
process or the same process. You create a Mach message port and set it 
up so that one thread will communicate to the exception thread through 
that port. So what you have in the B2 case is that I have created one 
thread that sits in a loop and all it does is wake-up whenever it hears 
about an exception and handles it. When it is awake the other thread 
that caused the exception is suspended. It is only that one thread that 
you setup to communicate with the exception handler that will 
communicate through the Mach port, which is what you want but different 
from how the signal code on a typical Unix works.

I should have some time tomorrow to look at the SS code, so I will take 
a look and see what I can do.

mzs

Re: [B2-devel] Re: BeOS Basilisk Broken in CVS

[Rob S. <ro...@xb...>]

Yeah I was not sure if the error was because of all the user modified 
BeOS's out there.  Since the fall of Be inc. , thing s have just gone 
nuts.   I did not know weather this was an issue becauased of updated 
video drivers (unified Nvidia Drivers) or the P4 optimized Kernel or what. 

What I would like to see is the rebirth of Be, with things like having 
the updated X86 SheepShaver on Be also getting the Latest BII working.   
Unfornately I understand very little  C++ and more of a hardware man my 
self.  I would like to help out any way possible with out having to 
write a line of code as I know there is nothing I can contribute.  While 
I can follow it to a degree I am no where on the level where I would 
need to be to suggest ideas.   There happens to a fellow on beshare name 
Scanity that who is very knowledgable and was talking about how 
possiblity one could improve sheepshaver .   I am going to try to hook 
him into to just look at the source and see what he thinks.


Nigel Pearson wrote:

>     Hi Rob.
>
>> video_beos.cpp:208: `VideoModes' undeclared (first use this function)
>> video_beos.cpp:208: (Each undeclared identifier is reported only once
>> video_beos.cpp:208: for each function it appears in.)
>> video_beos.cpp:209: implicit declaration of function `int
>> video_init_depth_list(...)'
>> video_beos.cpp:210: `VideoMonitor' undeclared (first use this function)
>> make: *** [obj.x86/video_beos.o] Error 1
>
>
>     In March 2002, the video interface changed. It became a C++ abstract
> base class, with the implementation in each platform's code area.
> It looks like the BeOS code still uses the old API.
>
> -- 
> Nigel Pearson, ni...@in... | "People say I'm strange, does it
> Telstra BI&D, Sydney, Australia       |        make me a stranger?
> Office: 8255 4222    Fax:  8255 3153  |  My best friend was born
> Mobile: 0408 664435  Home: 9792 6998  |         in a manger"    -DC Talk
>
>
>
> -------------------------------------------------------
> The SF.Net email is sponsored by EclipseCon 2004
> Premiere Conference on Open Tools Development and Integration
> See the breadth of Eclipse activity. February 3-5 in Anaheim, CA.
> http://www.eclipsecon.org/osdn
> _______________________________________________
> basilisk-devel mailing list
> bas...@li...
> https://lists.sourceforge.net/lists/listinfo/basilisk-devel
>

[B2-devel] Re: BeOS Basilisk Broken in CVS

[Nigel P. <ni...@in...>]

	Hi Rob.

> video_beos.cpp:208: `VideoModes' undeclared (first use this function)
> video_beos.cpp:208: (Each undeclared identifier is reported only once
> video_beos.cpp:208: for each function it appears in.)
> video_beos.cpp:209: implicit declaration of function `int
> video_init_depth_list(...)'
> video_beos.cpp:210: `VideoMonitor' undeclared (first use this function)
> make: *** [obj.x86/video_beos.o] Error 1

	In March 2002, the video interface changed. It became a C++ abstract
base class, with the implementation in each platform's code area.
It looks like the BeOS code still uses the old API.

--
Nigel Pearson, ni...@in... | "People say I'm strange, does it
Telstra BI&D, Sydney, Australia       |        make me a stranger?
Office: 8255 4222    Fax:  8255 3153  |  My best friend was born
Mobile: 0408 664435  Home: 9792 6998  |         in a manger"    -DC Talk

[B2-devel] BeOS Basilisk Broken in CVS

[Rob S. <ro...@xb...>]

Just a little introduction of my self since this is my first time 
posting to this list.
I am Robert Snyder aka Jawa78 on delphiforums and as far as I can tell 
the only maintainer of a Slackware install package of SheepShaver x86 as 
I like to call it.   During my travels of the web I have come across 
some projects that has brought great joy back into my wondering OS 
life.  Most people know me a slackware zealot as  of my great love of 
the Distrobution.  But about the same time another OS BeOS to be extact 
had very excited about this new Desktop OS (Beos 3)  Well Hardware 
progressed Be inc died,  leaving me up poop creek until I came across 
BeOS 5 Max V3 (which was nice but had way to many bugs to be used even 
on a rare basis and BeOS Dev Edition ( www.beosonline.com)  Which is 
more to my suting I had re entere the world of be to help find a 
developer that might take a shot at retrofitting as I call it all the 
great features that Gwenole ( sorry if I mis-spelled your name)  had  
put into the *nix versions.  But I would first was looking to build BII 
from nothing, so I download the BII and SheepShaver from the cvs just to 
have the source and to build BII. 

I followed the instructions  in the install file of the base Basilisk II 
directory go into the terminal and navigate to the BeOS directory.
 typed make and this is what i got . Note there was abunch of other 
stuff but this is the only thing that caused it to bomb out andexit the 
install. 


video_beos.cpp:208: `VideoModes' undeclared (first use this function)
video_beos.cpp:208: (Each undeclared identifier is reported only once
video_beos.cpp:208: for each function it appears in.)
video_beos.cpp:209: implicit declaration of function `int 
video_init_depth_list(...)'
video_beos.cpp:210: `VideoMonitor' undeclared (first use this function)
make: *** [obj.x86/video_beos.o] Error 1




If someone can explain what is going on here it would be great.  Thanks

Robert Snyder

Re: [B2-devel] Darwin port

[Gwenole B. <gb...@di...>]

> Actually, the sigsegv handler code is the same as for B2. I have just 
> hacked around to use Mach exceptions to handle screen faults (early so 
> that we don't need the whole context) + other faults like writes to 
> ROM. However, SheepShaver hangs early after a few seconds. I think 
> this is because we have not taught our Mach exception handler to 
> execute with an alternate stack (set by sigaltstack() or another means 
> we can define) as we should. Is that possible to do?

Hmm, this doesn't change anything.

Anyway, I have attached the current code I am experimenting. If 
MACH_EXCEPTIONS is set to 1 in main_unix.cpp, then MacOS would hang 
afterwards. I currently have no glue. Otherwise, it's "fine". i.e. does 
not hang but is slow.

Re: [B2-devel] Darwin port

[Gwenole B. <gb...@di...>]

Hi,

> I have not looked at the SS code but maybe the concepts are similar to 
> B2. Yes signals are crufty and slow on Darwin. What is more of a 
> problem is that how the register state is saved is different from 
> version to version.

But I think they placed enough padding to accomodate future changes. 
Hence, a context is generally more than 1 KB.

> I do not know what you mean exactly "exception filters" in this 
> context. It looks to me like you want to to be able to run some code 
> whenever you encounter a bad memory access (for vosf), illegal 
> instruction (probably for emulation purposes), and some sort of 
> communication (the SIGUSR2, BTW what is that for).

SIGUSR2 are used to trigger interrupts. i.e. there is a tick thread 
that periodically raises a SIGUSR2.

> This is relatively straight forward. The first two would be handled by 
> the exception handler, the third by a mach message. The message is 
> usually done with MIG but since it looks like you are using C++ and I 
> assume the code is organized similarly to B2 it would probably be 
> easier to shoehorn it directly in like I did with the B2 code.

Actually, the sigsegv handler code is the same as for B2. I have just 
hacked around to use Mach exceptions to handle screen faults (early so 
that we don't need the whole context) + other faults like writes to 
ROM. However, SheepShaver hangs early after a few seconds. I think this 
is because we have not taught our Mach exception handler to execute 
with an alternate stack (set by sigaltstack() or another means we can 
define) as we should. Is that possible to do?

Thanks,
Gwenole.

Re: [B2-devel] Darwin port

[Gwenole B. <gb...@di...>]

Hi,

> I co'ed a copy from CVS. The configure.in seems to use the __PAGEZERO 
> stuff from B2 but some parts do not appear to be in the repository.

You have to update your links. i.e. make links in SheepShaver/ 
directory. If B2 is not located into ../BasiliskII/ you can make links 
B2_TOPDIR=<whereever>/BasiliskII

Bye,
Gwenole.

Re: [B2-devel] Darwin port

[Michael S. <msl...@co...>]

Gwenole,

I co'ed a copy from CVS. The configure.in seems to use the __PAGEZERO 
stuff from B2 but some parts do not appear to be in the repository. 
Maybe you forgot to commit files like Darwin/testlmem.sh and 
Darwin/lowmem.c? main_unix.cpp seems to check for PAGEZERO_HACK or 
maybe you just have that stuff hard-coded?

mzs

Re: [B2-devel] Darwin port

[Michael S. <msl...@co...>]

On Jan 18, 2004, at 12:27 PM, Gwenole Beauchesne wrote:

> As you can see, video performance is crappy. Probably that using the 
> Mach exception filters will help a little instead of going through 
> signal handlers with full register state save. Does any Mach expert 
> know how to add exception filters for SIGUSR2/SIGILL things too?

I suppose I am the Mach expert on this mailing list.

I have not looked at the SS code but maybe the concepts are similar to 
B2. Yes signals are crufty and slow on Darwin. What is more of a 
problem is that how the register state is saved is different from 
version to version.

I do not know what you mean exactly "exception filters" in this 
context. It looks to me like you want to to be able to run some code 
whenever you encounter a bad memory access (for vosf), illegal 
instruction (probably for emulation purposes), and some sort of 
communication (the SIGUSR2, BTW what is that for). This is relatively 
straight forward. The first two would be handled by the exception 
handler, the third by a mach message. The message is usually done with 
MIG but since it looks like you are using C++ and I assume the code is 
organized similarly to B2 it would probably be easier to shoehorn it 
directly in like I did with the B2 code.

Tell me which files to look at specifically and give some details as to 
why you did things certain ways which will not be clear from just 
looking at the code and I will look into it and help you out.

mzs

Re: [B2-devel] Darwin port

[Gwenole B. <gb...@di...>]

Hi,

Darwin bits are now committed. Tested on MacOS X 10.2 with Apple X11  
server. Note you may notice that SheepShaver doesn't fully exit on  
shutdown for unknown reason at the moment.

> The Darwin port is almost working (to be committed):
> <http://gwenole.beauchesne.free.fr/sheepshaver/ 
> capture_darwin_port2.pdf>

BTW, you are better off to configure --disable-vosf for now there. It's  
better but not yet optimal either.

Bye,
Gwenole.