Commit Graph

1475 Commits

Author SHA1 Message Date
gbeauche
0f358518df New paranoia checks that better match was SheepShaver is doing with threads
and sigaltstack(). At least, this time we force use of threads and also
alter r1 to see whether the current threading model relies on it or not.
2005-07-04 06:09:59 +00:00
gbeauche
29207d9cf2 Move sigregs handling to dedicated header 2005-07-04 05:34:33 +00:00
gbeauche
d80a943daf Don't use sem_t definition from our local "semaphore.h" layer. Copy the
definitions from original LinuxThreads.
2005-07-03 23:28:30 +00:00
gbeauche
811b8d6648 Bad MacOS X, we need our our semaphore.h from Basilisk II to get POSIX
unnamed semaphores emulation with named Mach semaphores.
2005-07-03 22:44:30 +00:00
gbeauche
97eaae096c Fix typo for AO_transmit_packet() thunk. 2005-07-03 22:43:11 +00:00
gbeauche
e259aeb504 Tweaks to build slirp on MacOS X. Yeah, there is now ethernet there! ;-) 2005-07-03 22:42:48 +00:00
gbeauche
f346839f86 Add the complete NDRV variant for DIRECT_ADDRESSING modes. We could use
this one for all cases but I'd prefer keep it that way. i.e. the old
driver in REAL_ADDRESSING mode (with the D(bug()) facility), and the new
NDRV for DIRECT_ADDRESSING mode (e.g. Windows).
2005-07-03 22:04:18 +00:00
gbeauche
97726bd1e7 Minor tweaks to support compilation of ether.cpp within MacOS. i.e. mostly
migrate the Ethernet driver to the MacOS side. This is enabled for
DIRECT_ADDRESSING cases. I didn't want to alter much of ether.cpp (as it
would have required to support that mode). Of course, in REAL_ADDRESSING
mode (the default) and for debugging purposes, the old driver is still
available.
2005-07-03 22:02:01 +00:00
gbeauche
9c16ab1514 Remove obsolete and broken Cygwin/X11 hacks. Forbid builds of the Windows
version from within the Unix/ directory.
2005-07-03 13:39:06 +00:00
gbeauche
b7004b1199 Add necessary configury + support code to support slirp in SheepShaver,
the user-space network emulation layer. Enable it on all Unix supported
platforms where I know it works.
2005-07-03 13:05:36 +00:00
gbeauche
2cf7fd2a88 First round of patches to increase ethernet code portability. 2005-07-03 08:23:42 +00:00
gbeauche
9d7e5017d9 Factor out stuff for SheepShaver compatibility. 2005-07-03 08:21:32 +00:00
gbeauche
029d86ce4c Fix build with CodeWarrior in MacOS (9 under SheepShaver ;-) 2005-07-02 22:54:07 +00:00
gbeauche
7cc1bbc7b8 Issue a SysError(dsOldSystem) if we are trying to use MacOS < 8.1.0 with a
NewWorld ROM. That may be 8.1.0 included but original iMac had a NewWorld
ROM compatible system.

Otherwise we will crash because the boot routine is trying to execute code
through unitialized descriptor that points to 0x13ff, which is obviously
wrong (and unaligned on word-boundaries for 68k code).
2005-07-02 17:51:43 +00:00
gbeauche
2881c20813 Hopefully fix the remaining issue in the High Resolution Timing support
code and re-enable it on Linux platforms (they have clock_nanosleep). Why
did I trigger an interrupt inside a held lock? Hmmm, we should probably
add an _ack semaphore like we do e.g. for ethernet.
2005-07-01 23:15:11 +00:00
gbeauche
245c691102 Check for <stdint.h> ourselves + build with -mdynamic-no-pic on MacOS X. 2005-07-01 06:16:00 +00:00
gbeauche
f1ef1322b3 Add B2 Unix/m4 symlink 2005-07-01 06:15:21 +00:00
gbeauche
858c3b9719 - Check for <stdint.h> ourselves if AC_HEADER_STDC didn't do that
- Build with -mdynamic-no-pic on MacOS X, this brings up to 20% performance
  improvement in the CPU core.
2005-07-01 05:57:07 +00:00
gbeauche
01b34da3d9 Don't remove config.h.in in distclean rule if we are still keeping configure 2005-07-01 05:54:58 +00:00
gbeauche
eaa0f6576b Import m4 macros for external packages in case one would want to regenerate
a configure script without having those installed (e.g. on MacOS X).
2005-07-01 05:54:20 +00:00
gbeauche
a60b0fb78c Restore clobbered registers before calling SDL_PumpEvents(). 2005-06-30 22:26:12 +00:00
gbeauche
df960a814e HFS+ is case insensitive (sic!). At least, it pointed out to not remove
ppc_asm.s file but rather the temporary ppc_asm.out.s variant.
2005-06-30 22:19:01 +00:00
cebix
57609bf2fc modernized spec file and makefile 2005-06-30 21:27:26 +00:00
cebix
fa5ae3ab6e minor cleanups 2005-06-30 21:25:26 +00:00
gbeauche
ec8661f431 Fix a damn silly bug. On some occasions, we could have spcflags() set to
EXEC_RETURN | HANDLE_INTERRUPT. And then, we handled the interrupt, but
EXEC_RETURN was set so we returned very quickly without completing the
interrupt routine. As a side effect, this occasionnaly hung the emulator
most likely with {ethernet,audio}-based applications that trigger a lot
of interrupts.

The fix is to always honour EXEC_RETURN flag at first, of course.
2005-06-30 15:29:11 +00:00
gbeauche
fd69df24f3 Re-enable spinlocks on {i386,x86_64} since they are now used only for really
small atomic operations (add/sub). This implementation should be enough for
that purpose.
2005-06-30 13:29:43 +00:00
gbeauche
e867749999 POSIX conforming use of pthread_cond_wait(). 2005-06-30 10:34:31 +00:00
gbeauche
e355c83f59 Add system-specific implementations of idle_{wait,resume} functions. 2005-06-30 10:20:18 +00:00
gbeauche
08c5f8b713 Improve idle wait mechanism. Now, the emulator thread can be suspended
(idle_wait) until events arrived and notified through TriggerInterrupt().
i.e. we no longer sleep a fixed amount of time on platforms that support
a thread wait/signal mechanism.
2005-06-30 10:17:58 +00:00
gbeauche
c9b044aeaf Completely avoid any form of nested interrupt processing. 2005-06-30 09:09:59 +00:00
gbeauche
55f5e3a41b Fix CR save/restore in EmulOp thunk. I don't know what it fixes for real
but that was definitely wrong to only preserve CR2 there.
2005-06-30 07:34:17 +00:00
gbeauche
ea42af0568 Consume the packet even if we could not allocate a message block for it in
the Ethernet interrupt. The BeOS version does that, likewise for MOL. Otherwise,
we end up into an infinite loop reporting the memory allocation failure.

I think this is now the expected behavior as we wouldn't have stats
(num_rx_no_mem) for it if we couldn't get out of the EtherIRQ. ;-) Besides,
the packet will be resent for reliable networks.
2005-06-30 06:38:09 +00:00
gbeauche
7ac6ad2f46 Clean-ups: comments, ticks per sec output, don't restore r13 twice. 2005-06-28 16:50:30 +00:00
gbeauche
38d03d02e9 The alternate stack trick never worked as you can't modify an active stack.
i.e. it returned EPERM and ran into stack corruption to eventually crash the
emulator. This is noticeable in !hw_mac_cursor_accl mode (e.g. fullscreen DGA).

In order to the sigalstack() to be effective, we must kludge the kernel to
think it's running on another stack. In practise, we provide another stack
for the SIGUSR2 handler. sigusr2_handler_init() fulfills that purpose.

I hope this fixes remaining issues forever. At some point, I had multiple
*_init() handlers in case this is necessary.
2005-06-28 16:47:32 +00:00
gbeauche
a8ba24aecf Use fast spinlocks only for small enough atomic operations. Otherwise, you
run into some performance problems in e.g. video graphics experience because
of busywaits in the current spin_lock() implementation.
2005-06-25 11:40:29 +00:00
gbeauche
198925c5ae clean-ups from previous experiment with sched_yield(), this one caused a
slow-down too.
2005-06-25 11:36:35 +00:00
gbeauche
52138d2014 We HAVE_PTHREADS even if we use our own pthreads implementation, this also
induces availability of locking primitives. I will merge the !HAVE_PTHREADS
case (a la Basilisk II) for EMULATED_PPC when I get back to home.
2005-06-25 11:06:24 +00:00
gbeauche
b92ce82a32 Rewrite SheepThreads locks & semaphores. They now look much better and avoid
busy waits for acquiring spin locks.
2005-06-25 11:00:30 +00:00
gbeauche
e991e59fc1 fix test_and_set, aka fixes X async replies when clipboard active 2005-06-25 10:58:47 +00:00
gbeauche
2de44f1910 Fix for MacOS X assembler, why don't Apple use GNU as? 2005-06-25 06:33:39 +00:00
gbeauche
d0da2a0dda - Restore TOC on EMUL_OP_EXEC_RETURN from Execute68k
- Restore r13 on EMUL_RETURN, though I never saw r13 clobbered
- Increase stack frame allocated to EmulOp processing on par with BeOS
- Factor GetResource wrappers with a macro + restore r13 too
2005-06-24 22:58:55 +00:00
gbeauche
d8a33b0952 Direct Color visual is OK for fbdev dga. 2005-06-24 22:31:28 +00:00
gbeauche
26ee263b6f Don't fake the TVECT value on non-BeOS native systems. This is important
for systems that use a global r2 as the TLS register, e.g. Linux/ppc with
newer glibc. Also remove the syscall junk which were simply workarounds
for this bug. Remove a duplicate r2 restoration in EmulOp.

BTW, it's possible to get SheepShaver running on Linux/ppc systems with
NPTL rather than SheepThreads.
2005-06-23 16:23:31 +00:00
gbeauche
b9b90fce11 fix last commit 2005-06-23 11:56:43 +00:00
gbeauche
2deca51ca4 Add code to gather some stats on register usage. 2005-06-23 11:37:01 +00:00
gbeauche
c33f09b2d8 Fix crash in FBDev DGA mode when it failed to initialize (here, that was
because fb visual was Direct Color instead of True Color). We used to
initialize the_buffer* even if !open_display. Report that gracefully with
the dedicated STR_OPEN_WINDOW_ERR.
2005-06-22 16:40:24 +00:00
gbeauche
c41f403f56 Rework sheepshaver_cpu object allocation and get rid of POSIX'ish functions. 2005-06-22 16:38:15 +00:00
gbeauche
c4a7b10650 op_invoke*() symbols are expected to be in demangled form, i.e. "C" linkage 2005-06-22 16:35:59 +00:00
gbeauche
f887fb7bac Enable build of the JIT with GCC 2.9x compilers (e.g. "2.96" from MDK 8.1). 2005-06-22 15:37:25 +00:00
gbeauche
ae52321fe7 HOST_FLOAT_WORDS_BIG_ENDIAN is a tristate and, when defined, can still be
0, so little endian format. i.e. don't check that with a simple #ifdef...
2005-06-22 15:33:22 +00:00