Snes9x.com

BearOso wrote:

Screwtape wrote: Doing a little more testing, I have noticed that when the problem occurs, my Xorg.0.log file fills up with "mieqEnequeue: out-of-order valuator event; dropping.". A quick google reveals that this happens sometimes to people running some recent X.org versions - regardless of video-card or Linux distribution.

Thanks for looking into it for me, anyway.

I'm glad the problem isn't on my end, because unfortunately I have little recourse to fix it. Nonetheless, I've posted a revision 42 with a fix to the issue I was thinking was the cause, and might somehow be related. I hope you manage to solve the problem on your end.

Screwtape wrote: An update!

I switched to nouveau the other day, and tried snes9x-gtk again. This time XVideo works smooth as butter... except that everything is shaded blue - it looks like it's setting the blue component of every pixel to 255 and leaving the rest alone. Ah well.

BearOso wrote:*edit *edit** Another new version is up that just implements the feature you wanted without being done in a roundabout way.

BearOso wrote:I'm actually working on that very issue. The Xv color conversion code is severely broken. You can expect a fix in a day or so.

BearOso wrote:The revision 46 I promised is posted. This fixes the Xv driver to use the color information Xv provides.

Screwtape, you might need to use the newly-added inverted ordering option to get correct colors. I'm testing with the radeon driver on PPC, and X claims to want system byte-ordering, but for Xv the graphics card seems to need little-endian.

deniseweird wrote:There are some strange colours when I have the NTSC filter enabled now. I don't remember it being like this a version or two ago. I attached a screenshot:

If you look at the numerals here you should be able to see what I mean. Maybe this is just standard behaviour, but I don't remember it being like this before.

terran wrote:thanks for the screen saver fixes bearoso

One more request...do you know if there's been a change in the gtk code which handles the joystick configuration menu? I have a setup that doesn't have a mouse. I used to be able to use the tab-key to switch between button configuration boxes, but now if I hit tab, it ends up mapping the gamepad button to "tab." Not quite sure if there's a better UI flow...possible suggestions could be:
1) don't start grabbing input device output until you trigger with an "enter" key or something.
2) Make "ctrl-tab" or something flip between input-config boxes.
3) Disallow using tab as a key.

3 is not super ideal. 2 has kinda bad usability, but 1 probably takes the most work.

tradeoffs, tradeoffs. :-/

zerojay wrote:I'm having some trouble launching ROMs in snes9x-gtk from the command line. If I attempt to do so, the SNES9x window pops up for a splitsecond and disappears without any error.

Attempting to run a debug build through GDB to get a backtrace just outputs "program exited normally".

I'm running this on my PS3 with Yellow Dog Linux 6 installed. Both the precompiled packages for the distro as well as compiling from source exhibits this behavior. I used --with-joystick --without-opengl --without-assembler as my ./configure flags.

SNES9X-GTK works perfectly running the exact same ROM if I start up the GUI first... but I can't use the GUI for this since I'm going to be using a frontend.

(I can post more details such as config.log or anything else needed once I get home from work tonight, if needed. Yes, I am using version 49.)

Any ideas?

BearOso wrote:

zerojay wrote:I'm having some trouble launching ROMs in snes9x-gtk from the command line. If I attempt to do so, the SNES9x window pops up for a splitsecond and disappears without any error.

Attempting to run a debug build through GDB to get a backtrace just outputs "program exited normally".

I'm running this on my PS3 with Yellow Dog Linux 6 installed. Both the precompiled packages for the distro as well as compiling from source exhibits this behavior. I used --with-joystick --without-opengl --without-assembler as my ./configure flags.

SNES9X-GTK works perfectly running the exact same ROM if I start up the GUI first... but I can't use the GUI for this since I'm going to be using a frontend.

(I can post more details such as config.log or anything else needed once I get home from work tonight, if needed. Yes, I am using version 49.)

Any ideas?

Two questions:
First, does the ROM filename have spaces in it? The Snes9x core will not pull ROM names with spaces off the command-line correctly, but the GUI knows how to handle them. If this is the case, you can just rename the ROM and associated errata. Using alphanumerics only works best.

Second, are you trying to use any other options? Invalid options will cause the command-line parser to die and print a usage message, and this port doesn't provide most of the original unix options.

zerojay wrote: Yes, pretty sure there are spaces in the ROM names. Nope, no other options being set on the command line. I'll try it out tonight and let you know.

What I don't understand is why plain 1.51 SNES9x can handle ROMs with spaces in them without an issue, but the GTK version can't.

And by the way, thank you so much for your work on this port. It's incredible.

$ gdb ./snes9x-gtk
(gdb) break S9xExit
(gdb) break exit
Make breakpoint pending on future shared library load? (y or [n]) y
(gdb) run "/path/to/rom/here.smc"

(gdb) bt

zerojay wrote:I've removed the spaces from the zip files (leaving the spaces intact on the roms themselves) and it doesn't work.

I'll try to post a stack trace in a few minutes.

EDIT: Here's using a zip file with underscores replacing spaces, but the rom inside still using spaces.

Code: Select all

JACK tmpdir identified as [/dev/shm]
Starting sound
    --> (ALSA : default)...OK
Using ROM Super Mario Kart (USA).sfc in /home/jay/mount/ROMs/SNES/Super_Mario_Kart_(USA).zip
Map_HiROMMap
[New Thread 4137481392 (LWP 4592)]
[Thread 4137481392 (LWP 4592) exited]
[New Thread 4137481392 (LWP 4593)]
[Switching to Thread 4154573408 (LWP 4586)]

Breakpoint 1, S9xExit () at gtk/gtk_s9x.cpp:713
713         gui_config->save_config_file ();
(gdb) bt
#0  S9xExit () at gtk/gtk_s9x.cpp:713
#1  0x102386b4 in S9xHandlePortCommand (cmd=Cannot access memory at address 0x6
) at gtk/gtk_control.cpp:331
#2  0x100b72d0 in S9xApplyCommand (cmd=
        {type = 251 '?', multi_press = 0 '\0', button_norpt = 0 '\0', {button = {joypad = {idx = 0 '\0', toggle = 0 '\0', turbo = 0 '\0', sticky = 1 '\001', buttons = 0}, mouse = {idx = 0 '\0', left = 0 '\0', right = 0 '\0'}, scope = {fire = 0 '\0', cursor = 0 '\0', turbo = 0 '\0', pause = 0 '\0', aim_offscreen = 0 '\0'}, pointer = {idx = 0 '\0', speed_type = 0 '\0', UD = -1 '?', LR = 0 '\0'}, justifier = {idx = 0 '\0', trigger = 0 '\0', start = 0 '\0', aim_offscreen = 0 '\0'}, multi_idx = 100663296, command = 1536}, axis = {joypad = {idx = 0 '\0', invert = 0 '\0', axis = 3 '\003', threshold = 0 '\0'}, pointer = {idx = 0 '\0', speed_type = 0 '\0', invert = 1 '\001', HV = 1 '\001'}, button = {threshold = 6 '\006', negbutton = 0 '\0', posbutton = 0 '\0'}}, pointer = {aim_mouse0 = 0, aim_mouse1 = 0, aim_scope = 0, aim_justifier0 = 0, aim_justifier1 = 0}, port = "\006\000\000"}}, data1=0, data2=0) at controls.cpp:2045
#3  0x100b6e00 in S9xReportButton (id=553648130, pressed=false) at controls.cpp:1298
#4  0x10238f8c in S9xProcessEvents (block=<value optimized out>) at gtk/gtk_control.cpp:416
#5  0x1022fdac in S9xIdleFunc (data=<value optimized out>) at gtk/gtk_s9x.cpp:483
#6  0x0ec55158 in g_source_is_destroyed () from /lib/libglib-2.0.so.0
#7  0x0ec57914 in g_main_context_dispatch () from /lib/libglib-2.0.so.0
#8  0x0ec5b61c in g_main_context_check () from /lib/libglib-2.0.so.0
#9  0x0ec5ba84 in g_main_loop_run () from /lib/libglib-2.0.so.0
#10 0xf7cf6734 in gtk_main () from /usr/lib/libgtk-x11-2.0.so.0
#11 0x102303a8 in main (argc=2, argv=0xffc84764) at gtk/gtk_s9x.cpp:284
(gdb)                                                                    

I'll try it again with the zip and rom within having spaces removed soon.

EDIT: Just tried it with a ROM and ZIP that have all spaces removed. Didn't work. Also tried it with an unzipped ROM without spaces. Also does not work. On a whim, I thought that maybe the fact that my game is on a remote samba share could be causing the problem, but even when the ROM is local, it does not work.

I think we can say that loading ROMs from the command line is broken.