1 |
$Id: TODO,v 1.453 2007/02/19 01:34:42 debug Exp $ |
$Id: TODO,v 1.489 2007/05/01 04:05:06 debug Exp $ |
2 |
|
|
3 |
------------------------------------------------------------------------------- |
-------------------------------------------- |
4 |
|
|
5 |
Fix after the 0.4.4 release: |
Testing for the next release (0.4.5.1): |
|
Fix the PowerPC DECR interrupt speed! |
|
6 |
|
|
7 |
------------------------------------------------------------------------------- |
TEST DISK OVERLAY IMAGES |
8 |
|
TEST LANDISK/SUPERH EMULATION MODES |
9 |
|
REGRESSION TESTS FOR ALL OTHER SUPPORTED GUEST OSES |
10 |
|
|
11 |
|
# NetBSD/pmax 3.1 or 1.6.2 OK |
12 |
|
# NetBSD/arc 1.6.2 OK |
13 |
|
# NetBSD/hpcmips 3.1 OK |
14 |
|
# NetBSD/cobalt 3.1 OK |
15 |
|
# NetBSD/evbmips 3.1 OK |
16 |
|
# NetBSD/algor 3.1 OK |
17 |
|
# NetBSD/sgimips 3.1 OK |
18 |
|
# NetBSD/cats 3.1 OK |
19 |
|
# NetBSD/evbarm 2.1 OK |
20 |
|
# NetBSD/netwinder 3.1 OK |
21 |
|
# NetBSD/prep 2.1 OK |
22 |
|
# NetBSD/macppc 3.1 OK |
23 |
|
# NetBSD/dreamcast 3.1 MD OK |
24 |
|
# NetBSD/dreamcast 3.1 LiveCD OK |
25 |
|
# Linux/dreamcast Live CD OK |
26 |
|
# OpenBSD/pmax 2.8-BETA not tested because of lack of time |
27 |
|
# OpenBSD/cats 4.0 OK |
28 |
|
# OpenBSD/landisk 4.1 OK |
29 |
|
# Ultrix/RISC 4.5 OK |
30 |
|
# Sprite for DECstation OK |
31 |
|
# Debian GNU/Linux for pmax not tested because of lack of time |
32 |
|
|
33 |
Possible (relatively large) work packages to concentrate on in the future: |
Optional: |
34 |
|
# OpenBSD/sgi FAILED to boot after setup (as expected) |
35 |
|
|
36 |
x) SMP: |
-------------------------------------------- |
|
Get SMP working again. It is pretty much broken since I started |
|
|
the conversion from the old bintrans system to the new dyntrans system. |
|
|
Add better Test machine demos for SMP in the demos directory. |
|
37 |
|
|
38 |
x) Network: |
Some things, in totally random order, that I'd like to fix: |
39 |
Redesign of the networking subsystem, at least the NAT translation |
(Some items in this list are possibly out-of-date by now.) |
|
part. The current way of allowing raw ethernet frames to be |
|
|
transfered to/from the emulator via UDP should probably be extended |
|
|
to allow the frames to be transmitted other ways as well. |
|
|
Also adding support for connecting ttys (either to xterms, or to |
|
|
pipes/sockets etc, or even to PPP->NAT or SLIP->NAT :-). |
|
|
|
|
|
x) PCI: |
|
|
Pretty much everything related to runtime configuration, device |
|
|
slots, interrupts, whatever. The current code is very hardcoded |
|
|
and ugly. |
|
|
|
|
|
x) Debugging: |
|
|
Think more about SMP debugging, etc. Right now, the |
|
|
debugger is a mess. Also, a better connection to GDB would be |
|
|
very nice to have. |
|
|
|
|
|
x) Userland emulation: |
|
|
Primary goals would be NetBSD and Linux syscall emulation. |
|
|
|
|
|
And of course, there are _LOTS_ of minor TODOs spread out throughout |
|
|
the source code, which must be fixed sooner or later. |
|
|
|
|
|
------------------------------------------------------------------------------- |
|
|
|
|
|
Some other things, in random order, that I'd like to fix: (Some items in |
|
|
this list are probably out-to-date by now.) |
|
40 |
|
|
41 |
Dyntrans: |
Dyntrans: |
42 |
x) Instruction combination collisions? How to avoid easily... |
x) Instruction combination collisions? How to avoid easily... |
50 |
kludgy very fast.) |
kludgy very fast.) |
51 |
x) Dyntrans with SMP... lots of work to be done here. |
x) Dyntrans with SMP... lots of work to be done here. |
52 |
x) Dyntrans with cache emulation... lots of work here as well. |
x) Dyntrans with cache emulation... lots of work here as well. |
53 |
|
x) Remove the concept of base RAM completely; it would be more |
54 |
|
generic to allow RAM devices to be used "anywhere". |
55 |
o) dev_mp doesn't work well with dyntrans yet |
o) dev_mp doesn't work well with dyntrans yet |
56 |
o) In general, IPIs, CAS, LL/SC etc must be made to work with dyntrans |
o) In general, IPIs, CAS, LL/SC etc must be made to work with dyntrans |
57 |
x) Redesign/rethink the delay slot mechanism used for e.g. MIPS, |
x) Redesign/rethink the delay slot mechanism used for e.g. MIPS, |
72 |
(containing code) |
(containing code) |
73 |
x) 16-bit encodings? (MIPS16, ARM Thumb, 32-bit SH on SH64) |
x) 16-bit encodings? (MIPS16, ARM Thumb, 32-bit SH on SH64) |
74 |
x) Lots of other stuff: see src/cpus/README_DYNTRANS |
x) Lots of other stuff: see src/cpus/README_DYNTRANS |
75 |
x) true recompilation backend? think carefully about this. |
x) Native code generation backends: |
76 |
o) abstract syntax for emitting opcopdes |
o) think carefully about this. |
77 |
|
o) simple syntax for emitting opcodes; backend implementation |
78 |
|
must be optional, so I don't have to write more code |
79 |
|
than necessary. after all, the non-native (C) code should |
80 |
|
always work. |
81 |
o) convert into native code only after an entire |
o) convert into native code only after an entire |
82 |
block has been translated? probably best. |
block has been translated? probably best. |
83 |
|
o) the "almost native" opcodes may be rearranged, |
84 |
|
"peep-hole optimized", etc. and then as a separate step |
85 |
|
this list of almost native opcodes is written out |
86 |
|
as native code. |
87 |
|
o) think about delay slots at the end of a block! |
88 |
o) x86/amd64 code generator can be very similar... perhaps |
o) x86/amd64 code generator can be very similar... perhaps |
89 |
|
o) NOTE that generation is per _ABI_, not per host arch! |
90 |
|
the configure script must detect ABI!!! |
91 |
o) branches to already translated code blocks can |
o) branches to already translated code blocks can |
92 |
link the blocks together |
link the blocks together |
93 |
o) load/store are the most important. |
o) load/store are the most important to optimize |
94 |
|
|
95 |
Simple Valgrind-like checks? |
Simple Valgrind-like checks? |
96 |
o) Mark every address with bits which tell whether or not the address |
o) Mark every address with bits which tell whether or not the address |
143 |
register definitions according to http://techpubs.sgi.com/library/tpl/cgi-bin/getdoc.cgi/hdwr/bks/SGI_Developer/books/R10K_UM/sgi_html/t5.Ver.2.0.book_263.html#HEADING334 |
register definitions according to http://techpubs.sgi.com/library/tpl/cgi-bin/getdoc.cgi/hdwr/bks/SGI_Developer/books/R10K_UM/sgi_html/t5.Ver.2.0.book_263.html#HEADING334 |
144 |
and make sure everything works with R10000. |
and make sure everything works with R10000. |
145 |
Then test with OpenBSD/sgi? |
Then test with OpenBSD/sgi? |
146 |
|
x) Entry LO mask (as above). |
147 |
x) memory space, exceptions, ... |
x) memory space, exceptions, ... |
148 |
x) use cop0 framemask for tlb lookups |
x) use cop0 framemask for tlb lookups |
149 |
(http://techpubs.sgi.com/library/tpl/cgi-bin/getdoc.cgi/hdwr/bks/SGI_Developer/books/R10K_UM/sgi_html/t5.Ver.2.0.book_284.html) |
(http://techpubs.sgi.com/library/tpl/cgi-bin/getdoc.cgi/hdwr/bks/SGI_Developer/books/R10K_UM/sgi_html/t5.Ver.2.0.book_284.html) |
150 |
|
|
151 |
SuperH: |
SuperH: |
152 |
|
x) SH4 performance is VERY low when running user-space instructions, |
153 |
|
because I actually simulate the 4-entry ITLB as being separate |
154 |
|
from the 64-entry DTLB. This is correct, but VERY slow. I need |
155 |
|
to experiment with _not_ simulating it in too much detail. |
156 |
x) SH4 interrupt controller: |
x) SH4 interrupt controller: |
157 |
x) Implement correct priorities of interrupts |
x) MASKING should be possible! |
158 |
x) SH4 DMA (0xffa00000) |
x) SH4 DMA (0xffa00000) |
159 |
x) SH4 UBC (0xff200000) |
x) SH4 UBC (0xff200000) |
160 |
x) Store queues can copy 32 bytes at a time, there's no need to |
x) Store queues can copy 32 bytes at a time, there's no need to |
161 |
copy individual 32-bit words. (Performance improvement.) |
copy individual 32-bit words. (Performance improvement.) |
162 |
x) SH4 BSC (Bus State Controller) |
x) SH4 BSC (Bus State Controller) |
|
x) SH4 RTC: Read the host's clock. |
|
|
x) SH4 SCIF: Serial _interrupts_ |
|
163 |
x) Instruction tracing should include symbols for branch targets, |
x) Instruction tracing should include symbols for branch targets, |
164 |
and so on, to make the output more human readable. |
and so on, to make the output more human readable. |
165 |
x) NetBSD/evbsh3, dreamcast, mmeye, hpcsh! Linux? |
x) SH3-specific devices: Pretty much everything! |
166 |
|
x) NetBSD/evbsh3, mmeye, hpcsh! Linux? |
167 |
x) Replace pc-relative loads with immediate load, if within the |
x) Replace pc-relative loads with immediate load, if within the |
168 |
same page. (Similar to the same optimization for ARM.) |
same page. (Similar to the same optimization for ARM.) |
169 |
x) Floating point speed! |
x) Floating point speed! |
170 |
x) Floating point exception correctness. |
x) Floating point exception correctness. |
171 |
|
x) NetBSD HEAD (as of April 2007) hangs during bootup, because it |
172 |
|
turns on/off interrupts in an unfortunately synchronized way |
173 |
|
with dyntrans. This needs to be fixed. |
174 |
|
x) Exceptions for unaligned load/stores. OpenBSD/landisk uses |
175 |
|
this mechanism for its reboot code (machine_reset). |
176 |
x) Think carefully about how to implement SH5/SH64 (for evbsh5). |
x) Think carefully about how to implement SH5/SH64 (for evbsh5). |
177 |
|
|
178 |
|
Landisk SH4: |
179 |
|
x) When NetBSD/landisk 4.0 has been released, make sure it works |
180 |
|
in the emulator. (Update documentation, etc.) |
181 |
|
|
182 |
Dreamcast: |
Dreamcast: |
183 |
x) G2 DMA |
x) G2 DMA |
184 |
x) LAN adapter (dev_mb8696x.c). NetBSD root-on-nfs. |
x) LAN adapter (dev_mb8696x.c). NetBSD root-on-nfs. |
185 |
x) PVR: Lots of stuff. See dev_pvr.c. |
x) PVR: Lots of stuff. See dev_pvr.c. |
186 |
x) GDROM |
x) Better GDROM support |
187 |
x) Modem |
x) Modem |
188 |
x) PCI bridge/bus? |
x) PCI bridge/bus? |
189 |
x) Maple bus: |
x) Maple bus: |
193 |
x) GD-ROM emulation: Use the GDROM device. |
x) GD-ROM emulation: Use the GDROM device. |
194 |
x) Use the VGA font as a fake ROM font. (Better than |
x) Use the VGA font as a fake ROM font. (Better than |
195 |
nothing.) |
nothing.) |
|
x) Linux/dreamcast? (The gentoo kernel currently crashes.) |
|
196 |
x) Make as many as possible of the KOS examples run! |
x) Make as many as possible of the KOS examples run! |
197 |
x) More homebrew demos/games. |
x) More homebrew demos/games. |
198 |
x) SPU: Sound emulation (ARM cpu). |
x) SPU: Sound emulation (ARM cpu). |
201 |
http://www.maushammer.com/vmu.html for a good description of |
http://www.maushammer.com/vmu.html for a good description of |
202 |
the differences between LC86104C and the one used in the VME. |
the differences between LC86104C and the one used in the VME. |
203 |
|
|
|
Transputer: |
|
|
x) Implement support for Helios binaries. |
|
|
x) Stack and register contents at startup? |
|
|
x) Figure out how to boot an entire Helios distribution. |
|
|
x) Implement all instructions. :) |
|
|
|
|
|
RCA1802/RCA1805, CHIP8: |
|
|
x) CHIP8 -> RCA180x conversion |
|
|
x) Think about how to do dual-mode, variable-instr-length |
|
|
ISAs, and switch between modes. |
|
|
x) 1805 "extended" opcode -> trigger CHIP8 emulation? |
|
|
That is, all calls 0NNN could point to 0x68 opcodes, |
|
|
which, if running on a 1802 in CHIP8-emulation-mode, |
|
|
would be manually interpreted. |
|
|
x) Better solution: |
|
|
CHIP8 calls to 00xx => handle at high level, |
|
|
calls to 0xxx in general = call 180X machine code |
|
|
(0000 = reboot?) |
|
|
x) 1802 info: http://www.nyx.net/~lturner/public_html/Cosmac.html |
|
|
and: http://www.elf-emulation.com/1802.html |
|
|
x) 1805 extended opcodes: Implement at least disassembly support! |
|
|
x) Keyboard input. |
|
|
x) Sound (beep only). |
|
|
x) Slow-down to correct speed? Wikipedia: "it was usually operated |
|
|
at 3.58 MHz/2 to suit the requirements of the 1861 chip which |
|
|
gave a speed of a little over 100,000 instructions per second" |
|
|
(Note that _CHIP8_ emulation would then be even slower.) |
|
|
x) SCHIP48 (Super) emulation: |
|
|
Some more opcodes, 128x64 framebuffer, larger |
|
|
sprites and fonts. |
|
|
|
|
204 |
Alpha: |
Alpha: |
205 |
x) OSF1 PALcode, Virtual memory support. |
x) OSF1 PALcode, Virtual memory support. |
206 |
x) PALcode replacement! PAL1E etc opcodes...? |
x) PALcode replacement! PAL1E etc opcodes...? |
210 |
x) More Alpha machine types, so it could work with |
x) More Alpha machine types, so it could work with |
211 |
OpenBSD, FreeBSD, and Linux too? |
OpenBSD, FreeBSD, and Linux too? |
212 |
|
|
213 |
SPARC: |
SPARC (both the ISA and the machines): |
214 |
o) Implement Adress space identifiers; load/stores etc. |
o) Implement Adress space identifiers; load/stores etc. |
215 |
|
o) Exception/trap/interrupt handling. |
216 |
o) Save/restore register windows etc! Both v9 and pre-v9! |
o) Save/restore register windows etc! Both v9 and pre-v9! |
217 |
o) Finish the subcc and addcc flag computation code. |
o) Finish the subcc and addcc flag computation code. |
218 |
o) Add more registers (floating point, control regs etc) |
o) Add more registers (floating point, control regs etc) |
|
o) Exception/trap handling. |
|
219 |
o) Disassemly of some more instructions? |
o) Disassemly of some more instructions? |
220 |
o) Are sll etc 32-bit sign-extending or zero-extending? |
o) Are sll etc 32-bit sign-extending or zero-extending? |
221 |
o) Finish the GDB register stuff. |
o) Floating point exception correctness. |
|
x) Floating point exception correctness. |
|
222 |
o) SPARC v8, v7 etc? |
o) SPARC v8, v7 etc? |
223 |
|
o) More machine modes and devices. |
224 |
|
|
225 |
Debugger: |
Debugger: |
226 |
o) How does SMP debugging work? Does it simply use "threads"? |
o) How does SMP debugging work? Does it simply use "threads"? |
265 |
x) Alignment exceptions. |
x) Alignment exceptions. |
266 |
|
|
267 |
PReP: |
PReP: |
268 |
Clock time! ("Bad battery blah blah") |
x) Clock time! ("Bad battery blah blah") |
269 |
|
|
270 |
Algor: |
Algor: |
271 |
o) Other models than the P5064? |
o) Other models than the P5064? |
280 |
x) Mouse/pad support! :) |
x) Mouse/pad support! :) |
281 |
x) A NIC? (As a PCMCIA device?) |
x) A NIC? (As a PCMCIA device?) |
282 |
|
|
283 |
|
M88K: |
284 |
|
o) Everything. :) |
285 |
|
o) More instruction disassembly! |
286 |
|
o) Implement more instructions. |
287 |
|
o) has-delay-slot (for debugging) |
288 |
|
o) Find manuals! |
289 |
|
o) MMU stuff |
290 |
|
o) Exceptions |
291 |
|
o) FPU |
292 |
|
o) Control registers |
293 |
|
|
294 |
AVR: |
AVR: |
295 |
o) Everything. |
o) Everything. |
296 |
|
|
301 |
o) try to get netbsd/evbarm 3.x or 4.x running (iq80321) |
o) try to get netbsd/evbarm 3.x or 4.x running (iq80321) |
302 |
o) make the xscale counter registers (ccnt) work |
o) make the xscale counter registers (ccnt) work |
303 |
o) make the ata controller usable for FreeBSD! |
o) make the ata controller usable for FreeBSD! |
304 |
o) Zaurus emulation, for e.g. OpenBSD/zaurus |
o) Zaurus emulation: |
305 |
|
x) OpenBSD/zaurus |
306 |
|
x) NetBSD/zaurus? See the following URL: |
307 |
|
http://mail-index.netbsd.org/port-arm/2006/11/19/0000.html |
308 |
o) Debian/cats crashes because of unimplemented coproc stuff. |
o) Debian/cats crashes because of unimplemented coproc stuff. |
309 |
fix this? |
fix this? |
310 |
|
|
328 |
possibly other live-CD formats.) |
possibly other live-CD formats.) |
329 |
|
|
330 |
Networking: |
Networking: |
331 |
|
x) Redesign of the networking subsystem, at least the NAT translation |
332 |
|
part. The current way of allowing raw ethernet frames to be |
333 |
|
transfered to/from the emulator via UDP should probably be |
334 |
|
extended to allow the frames to be transmitted other ways as |
335 |
|
well. |
336 |
|
x) Also adding support for connecting ttys (either to xterms, or to |
337 |
|
pipes/sockets etc, or even to PPP->NAT or SLIP->NAT :-). |
338 |
|
x) Documentation updates (!) are very important, making it easier to |
339 |
|
use the (already existing) network emulation features. |
340 |
x) Fix performance problems caused by only allowing a |
x) Fix performance problems caused by only allowing a |
341 |
single TCP packet to be unacked. |
single TCP packet to be unacked. |
342 |
x) Don't hardcode offsets into packets! |
x) Don't hardcode offsets into packets! |
424 |
x) refactor various clocks/nvram/cmos into one device? |
x) refactor various clocks/nvram/cmos into one device? |
425 |
|
|
426 |
PCI: |
PCI: |
427 |
|
x) Pretty much everything related to runtime configuration, device |
428 |
|
slots, interrupts, etc must be redesigned/cleaned up. The current |
429 |
|
code is very hardcoded and ugly. |
430 |
|
o) Allow cards to be added/removed during runtime more easily. |
431 |
|
o) Allow cards to be enabled/disabled (i/o ports, etc, like |
432 |
|
NetBSD needs for disk controller detection). |
433 |
|
o) Allow devices to be moved in memory during runtime. |
434 |
|
o) Interrupts per PCI slot, etc. (A-D). |
435 |
|
o) PCI interrupt controller logic... very hard to get right, |
436 |
|
because these differ a lot from one machine to the next. |
437 |
x) last write was ffffffff ==> fix this, it should be used |
x) last write was ffffffff ==> fix this, it should be used |
438 |
together with a mask to get the correct bits. also, not ALL |
together with a mask to get the correct bits. also, not ALL |
439 |
bits are size bits! (lowest 4 vs lowest 2?) |
bits are size bits! (lowest 4 vs lowest 2?) |
441 |
x) generalize the interrupt routing stuff (lines etc) |
x) generalize the interrupt routing stuff (lines etc) |
442 |
|
|
443 |
Clocks and timers: |
Clocks and timers: |
444 |
|
x) Fix the PowerPC DECR interrupt speed! (MacPPC and PReP speed, etc.) |
445 |
x) DON'T HARDCODE 100 HZ IN cpu_mips_coproc.c! |
x) DON'T HARDCODE 100 HZ IN cpu_mips_coproc.c! |
446 |
x) Test the 8253? Right now it doesn't seem to be used? |
x) Test the 8253? Right now it doesn't seem to be used? |
447 |
x) NetWinder timeofday is incorrect! |
x) NetWinder timeofday is incorrect! It seems to be exactly |
448 |
|
1 day ahead of actual time? |
449 |
x) Cobalt TOD is incorrect! |
x) Cobalt TOD is incorrect! |
450 |
x) Go through all other machines, one by one, and fix them. |
x) Go through all other machines, one by one, and fix them. |
451 |
|
|
464 |
o) non-IEEE modes (i.e. x86)? |
o) non-IEEE modes (i.e. x86)? |
465 |
|
|
466 |
Userland emulation: |
Userland emulation: |
467 |
x) Lots of stuff; freebsd and netbsd (and linux?) syscalls. |
x) Dynamic linking! |
468 |
x) Dynamic linking? Hm. |
x) Lots of stuff; freebsd, netbsd, linux, ... syscalls. |
469 |
|
x) Initial register/stack contents (environment, command line args). |
470 |
|
x) Return value (from main). |
471 |
|
x) mmap emulation layer |
472 |
|
x) errno emulation layer |
473 |
|
x) struct conversions for may syscalls |
474 |
|
|
475 |
Sound: |
Sound: |
476 |
x) generic sound framework |
x) generic sound framework |