Vm-dev January 2006

vm-dev@lists.squeakfoundation.org

11 participants
11 discussions

Can plugins add libraries to unix vm?
by Alan Sibley 28 Mar '06

28 Mar '06

Hey folks, We are working on a plugin using the unix VM for squeak under OSX Tiger. I'd like to have the resulting squeak VM link in some foundation stuff from OSX (CoreFoundation CoreServices...). Is there a way to trigger this using the Makefile.in stuff (tried altering XLDFLAGS, PLIBS, LIBS with no success). It would be nice to have this linkage only occur if adding in our plugin. Thanks in advance, Alan

7 46

svn server dead?
by Wolfgang Helbig 25 Jan '06

25 Jan '06

Hi, I tried the command svn checkout http://squeak.hpl.hp.com/svn/squeak/trunk/platforms/ to checkout. But it didn't work. I found the above command in squeak swiki. This one worked fine however: svn checkout http://squeakvm.org/svn/squeak/trunk/platforms/ Is the former one dead? Not existing any more? If so, I'd volunteer to update the swiki page. Wolfgang

2 3

Trying to write a plugin - how to store a pointer
by Todd Blanchard 24 Jan '06

24 Jan '06

For various reasons, I'm wanting to build a plugin to a library that requires me to hold onto pointers to things. This means I need to store the pointer in an ivar in a squeak object and use it on subsequent calls. Can I get away with using a SmallInteger for this? Basically doing something like: primNewContext | ctx | self export: true. self primitive: 'primCreateContext' parameters: #(). self var: #ctx type: 'int'. ctx := self cCode: 'LIB_NewContext()'. ^ctx. primDestroyContext: ctx self export: true. self primitive: 'primDestroyContext' parameters: #(SmallInteger). self cCode: 'LIB_DestroyContext((void*)ctx)'. Or is this evil and I should consider a different technique? If this is evil, what is the good way to store external addresses in ivars? Thanks

4 7

Re: Sound on 3.7-7 for solaris 8 on intel
by Wolfgang Helbig 20 Jan '06

20 Jan '06

>Hi, > >after installing successfully the 3.7-7 stuff from squeakvm.org on my sol8-i386 >system, playing music using the vm-sound-Sun plugin is kind of disappointing: >If I hit the play button to play the example midi file it plays less then >a second. If I hit the play button again, nothing happens. > >So what can be done? I tried to apply NAS, but I didn't find a vm-sound-NAS >plugin for solaris 8 on intel. > >thanks for any help! > >Wolfgang Guess I had to help myself :-) Now both vm-sound-NAS and vm-sound-Sun work on solaris x86, but I had to patch around little bit the contents of Squeak-3.7-7.src.tar.gz - To make it compile: I had to get rid of MAKEFLAGS that are not understood by Sun's make. - To make configure enable NAS export CPPFLAGS='-I /usr/openwin/include' - To make sqUnixExternalPrims.c accept the NAS module, I had to make it lazy (dlopen flag RTLD_NOW replayced by RTLD_LAZY. - To make NAS sound nice, I had to swap bytes, in fact I had to lie to it and tell him, that our format is MSB, but it is LSB in fact. (I think x86 is little endian, isn't it?) Thats all to make NAS work with squeak. Now the vm-sound-Sun module nearly worked out of the box, but it only played two buffers (about 200 ms). The asynch callback didn't work. I had to replace the aioHandle flag AIO_RX by AIO_RWX. I prefer the Sun module. It has no delay as apposed the NAS module, that has a delay of about 3 seconds, which is kind of irritating when you watch an MPEG move. Anyway, squeak is a great thing ... Thanks to all, who made it happen! Wolfgang *** Squeak-3.7-7/platforms/unix/config/make.cfg.in Sat Mar 19 21:38:28 2005 --- Squeak-3.7-7/platforms/unix/config/make.cfg.in Tue Jan 17 08:07:45 2006 *************** *** 37,43 **** # # Last edited: 2003-08-31 19:55:21 by piumarta on emilia.inria.fr ! MAKEFLAGS= --no-print-directory @SET_MAKE@ --- 37,43 ---- # # Last edited: 2003-08-31 19:55:21 by piumarta on emilia.inria.fr ! MAKEFLAGS= @SET_MAKE@ *** Squeak-3.7-7/platforms/unix/vm/sqUnixExternalPrims.c Sat Mar 19 21:38:22 2005 --- Squeak-3.7-7/platforms/unix/vm/sqUnixExternalPrims.c Tue Jan 17 13:57:16 2006 *************** *** 143,149 **** else { dprintf((stderr, "tryLoading %s\n", libName)); ! handle= dlopen(libName, RTLD_NOW | RTLD_GLOBAL); if (handle == 0) { if ((!err) && !(sqIgnorePluginErrors)) --- 143,149 ---- else { dprintf((stderr, "tryLoading %s\n", libName)); ! handle= dlopen(libName, RTLD_LAZY | RTLD_GLOBAL); if (handle == 0) { if ((!err) && !(sqIgnorePluginErrors)) *** Squeak-3.7-7/platforms/unix/vm-sound-NAS/sqUnixSoundNAS.c Sat Mar 19 21:38:26 2005 --- Squeak-3.7-7/platforms/unix/vm-sound-NAS/sqUnixSoundNAS.c Fri Jan 20 03:23:24 2006 *************** *** 48,58 **** static void dprintf(char *fmt, ...) {} #endif ! #ifdef WORDS_BIGENDIAN - # define AU_FORMAT AuFormatLinearSigned16MSB - #else # define AU_FORMAT AuFormatLinearSigned16LSB #endif --- 48,58 ---- static void dprintf(char *fmt, ...) {} #endif ! /* NAS sounds better on Solaris x86 if you lie to it: */ #ifdef WORDS_BIGENDIAN # define AU_FORMAT AuFormatLinearSigned16LSB + #else + # define AU_FORMAT AuFormatLinearSigned16MSB #endif *************** *** 65,74 **** #define BYTES_PER_SAMPLE 2 /* Squeak always uses 16-bit samples */ ! #ifdef HAS_MSB_FIRST ! # define IS_BIGENDIAN 1 ! #else # define IS_BIGENDIAN 0 #endif --- 65,74 ---- #define BYTES_PER_SAMPLE 2 /* Squeak always uses 16-bit samples */ ! #ifdef LSB_FIRST # define IS_BIGENDIAN 0 + #else + # define IS_BIGENDIAN 1 #endif *************** *** 157,163 **** char *buf; /* buffer to play from; it may not be arrayIndex if a conversion is necessary */ ! dprintf("PlaySamples(frameCount=%d, arrayIndex=%d, startIndex=%d\n", frameCount, arrayIndex, startIndex); /* figure out how much to play */ --- 157,163 ---- char *buf; /* buffer to play from; it may not be arrayIndex if a conversion is necessary */ ! dprintf("PlaySamples(frameCount=%d, arrayIndex=%u, startIndex=%d\n", frameCount, arrayIndex, startIndex); /* figure out how much to play */ *************** *** 378,384 **** frameCount, frameCount / 4); AuMakeElementImportClient(&elements[0], samplesPerSec, ! AuFormatLinearSigned16LSB, /* XXX this should be chosen based on the platform */ stereo ? 2 : 1, AuTrue, 2*frameCount, /* max: 2 buffers */ --- 378,384 ---- frameCount, frameCount / 4); AuMakeElementImportClient(&elements[0], samplesPerSec, ! AU_FORMAT, stereo ? 2 : 1, AuTrue, 2*frameCount, /* max: 2 buffers */ *************** *** 484,490 **** AuMakeElementExportClient(&elements[1], 0, desiredSamplesPerSec, ! AuFormatLinearSigned16LSB, /* XXX this should be chosen based on the platform */ stereo ? 2 : 1, AuTrue, 1000000, /* was AuUnlimitedSamples */ --- 484,490 ---- AuMakeElementExportClient(&elements[1], 0, desiredSamplesPerSec, ! AU_FORMAT, stereo ? 2 : 1, AuTrue, 1000000, /* was AuUnlimitedSamples */ diff -rbc Squeak-3.7-7/platforms/unix/vm-sound-Sun/sqUnixSoundSun.c Squeak-3.7-7/platforms/unix/vm-sound-Sun/sqUnixSoundSun.c *** Squeak-3.7-7/platforms/unix/vm-sound-Sun/sqUnixSoundSun.c Sat Mar 19 21:38:51 2005 --- Squeak-3.7-7/platforms/unix/vm-sound-Sun/sqUnixSoundSun.c Fri Jan 20 03:06:41 2006 *************** *** 65,71 **** static int sound_Stop(void); ! static int auFd= -1; /* open on /dev/dsp */ static int fmtStereo= 0; /* whether we are playing in stereo or not */ static int auPlaySemaIndex= 0; /* an index to signal when new data may be played */ static int auBufBytes= 0; /* buffer size to use for playback. --- 65,71 ---- static int sound_Stop(void); ! static int auFd= -1; /* open on /dev/audio */ static int fmtStereo= 0; /* whether we are playing in stereo or not */ static int auPlaySemaIndex= 0; /* an index to signal when new data may be played */ static int auBufBytes= 0; /* buffer size to use for playback. *************** *** 121,127 **** perror("AUDIO_GETINFO"); goto closeAndFail; } ! info.play.gain= 100; info.play.precision= 16; info.play.encoding= AUDIO_ENCODING_LINEAR; info.play.channels= fmtStereo ? 2 : 1; --- 121,127 ---- perror("AUDIO_GETINFO"); goto closeAndFail; } ! info.play.gain= 255; /* this is 100 percent */ info.play.precision= 16; info.play.encoding= AUDIO_ENCODING_LINEAR; info.play.channels= fmtStereo ? 2 : 1; *************** *** 137,143 **** goto closeAndFail; } aioEnable(auFd, 0, 0); ! aioHandle(auFd, auHandle, AIO_RX); return true; closeAndFail: --- 137,143 ---- goto closeAndFail; } aioEnable(auFd, 0, 0); ! aioHandle(auFd, auHandle, AIO_RWX); return true; closeAndFail:

1 0

Large and small integers on 64 bit Squeak
by David T. Lewis 15 Jan '06

15 Jan '06

For 32 bit Squeak on 32 bit Intel hardware: SmallInteger maxVal hex ==> '16r3FFFFFFF' 16r3FFFFFFF class ==> SmallInteger 16r40000000 class ==> LargePositiveInteger 16r3FFFFFFF + 1 = 16r40000000 ==> true For 64 bit Squeak on 32 bit Intel hardware: SmallInteger maxVal hex ==> '16r3FFFFFFF' 16r4FFFFFF03FFFFFFF class ==> SmallInteger 16r4FFFFFF040000000 class ==> LargePositiveInteger (16r4FFFFFF03FFFFFFF + 1) = 16r4FFFFFF040000000 ==> true For 64 bit Squeak on 64 bit AMD hardware (note high order bit): SmallInteger maxVal hex ==> '16r3FFFFFFF' 16rFFFFFFF03FFFFFFF class ==> SmallInteger 16rFFFFFFF040000000 class ==> LargePositiveInteger (16rFFFFFFF03FFFFFFF + 1) = 16rFFFFFFF040000000 ==> true SmallIntegers are still 31 bits in 64 bit Squeak, so this does not look right. Or am I just misunderstanding something? I was looking at this because of DateAndTime now ==> -4391-03-28T17:46:11+00:00 on 64 bit Squeak, which definitely does not look right. Dave

1 1

Trouble building unix VM for OSX
by Alan Sibley 15 Jan '06

15 Jan '06

Hey folks, I'm trying to build the latest unix/osx from svn rev. 1290. I think I have the latest VMMaker (2.8b5-64B). I'm compiling but not linking... missing sqGetFilenameFromString. It's referenced from ioFilenamefromStringofLengthresolveAliases in interp.c. I grepped around and didn't find it... heeeeelp ;) Thanks in advance, Alan

3 2

Curious difference in behavior
by Todd Blanchard 14 Jan '06

14 Jan '06

I'm going back and forth between a Mac using the Unix VM and a PC using the windows one quite a lot with the same image. I recently coded myself an infinite loop by accident and found that the Windows VM has no trouble interrupting the execution and bringing up a debugger to find the problem (using alt period) but the Unix one, while it opens debugger windows when holding down cmd period, still remains frozen and I have to quit the VM. Does anyone know why this is different and, more importantly, how we might fix it?

2 1

Sound on 3.7-7 for solaris 8 on intel
by Wolfgang Helbig 14 Jan '06

14 Jan '06

Hi, after installing successfully the 3.7-7 stuff from squeakvm.org on my sol8-i386 system, playing music using the vm-sound-Sun plugin is kind of disappointing: If I hit the play button to play the example midi file it plays less then a second. If I hit the play button again, nothing happens. So what can be done? I tried to apply NAS, but I didn't find a vm-sound-NAS plugin for solaris 8 on intel. thanks for any help! Wolfgang

1 0

Smartphone port of squeak
by Sam Liddicott 13 Jan '06

13 Jan '06

I'm trying my hand at porting squeak to MS Smartphone 2003, which is based on WINCE4. I'm adapting the win32 squeak build, which makes the most sense; Smartphone, however, DOES have a joystick but does NOT have a mouse or mouse pointer. It does have a concept of tab-order and focus. Is this going to be a dead loss? while faking a mouse pointer might be worthwhile for some things, if smalltalk is useless without it, then it is probably a waste of time; does smalltalk or the GUI used, have a concept of focus? Or would these be a feature of something like wxSqueak, which probably could be ported? Comments? The other problem is, that although Smartphone supports sizeof(char)=1, 2 byte wchar_t is used for most API calls. Comments? Sam

2 1

Obscure hang
by Bert Freudenberg 09 Jan '06

09 Jan '06

Hi folks, I've had some rather obscure hangs on my squeak-source server lately. The funny thing is, as soon as I connect to the Xvnc server and move the mouse, things start rolling again. Otherwise, it takes a couple of minutes until serving continues. I have no idea why. Here are the file descriptors for reference. #3 is the X server (Xvnc) connection: squeaksou 30574 source 0r CHR 1,3 31583 /dev/ null squeaksou 30574 source 1w REG 3,5 32895 16482319 .../ some.log squeaksou 30574 source 2w REG 3,5 32895 16482319 .../ some.log squeaksou 30574 source 3u unix 0xf2119ac0 3367464 socket squeaksou 30574 source 4r REG 3,5 23206828 12402706 .../ some.image squeaksou 30574 source 5r DIR 3,5 4096 12402698 .../ somedir squeaksou 30574 source 6r REG 3,5 14542313 17317896 .../ SqueakV3.sources squeaksou 30574 source 7u REG 3,5 20230368 12402705 .../ some.changes squeaksou 30574 source 8u IPv4 4290750 TCP *:someport (LISTEN) Unfortunately I took this only after the hang. Socket #9 must be a file that was interrupted while writing, 10 and 11 is an HTTP connection pair. There were already 3581255 bytes of 4457385 bytes written to file #9 at this point. === According to strace the VM was chugging along just fine: gettimeofday({1136823729, 265939}, NULL) = 0 gettimeofday({1136823729, 265985}, NULL) = 0 gettimeofday({1136823729, 266050}, NULL) = 0 gettimeofday({1136823729, 266092}, NULL) = 0 --- SIGALRM (Alarm clock) @ 0 (0) --- sigreturn() = ? (mask now []) gettimeofday({1136823729, 277789}, NULL) = 0 gettimeofday({1136823729, 277819}, NULL) = 0 gettimeofday({1136823729, 277856}, NULL) = 0 gettimeofday({1136823729, 277877}, NULL) = 0 gettimeofday({1136823729, 277910}, NULL) = 0 gettimeofday({1136823729, 277939}, NULL) = 0 ioctl(3, FIONREAD, [0]) = 0 gettimeofday({1136823729, 278015}, NULL) = 0 select(9, [3], [], [3 8], {0, 0}) = 0 (Timeout) gettimeofday({1136823729, 278212}, NULL) = 0 gettimeofday({1136823729, 289711}, NULL) = 0 gettimeofday({1136823729, 289751}, NULL) = 0 gettimeofday({1136823729, 289813}, NULL) = 0 gettimeofday({1136823729, 289854}, NULL) = 0 ... === but nothing was served to sockets 10/11 nor written to file #9. Now some X data arrives on socket #3 ... gettimeofday({1136823748, 170924}, NULL) = 0 gettimeofday({1136823748, 170956}, NULL) = 0 gettimeofday({1136823748, 170984}, NULL) = 0 ioctl(3, FIONREAD, [32]) = 0 read(3, "\7\1W\204A\306\371\257%\0\0\0\1\0\200\0\2\0\200\0\217 \003"..., 32) = 32 ioctl(3, FIONREAD, [32]) = 0 read(3, "\6\0W\204B\306\371\257%\0\0\0\2\0\200\0\0\0\0\0\211\003"..., 32) = 32 gettimeofday({1136823748, 171177}, NULL) = 0 ioctl(3, FIONREAD, [0]) = 0 gettimeofday({1136823748, 171221}, NULL) = 0 select(9, [3], [], [3 8], {0, 0}) = 0 (Timeout) ... ==== and 5 seconds later it answers the (pending?) request on sockets 10/11 ioctl(3, FIONREAD, [0]) = 0 gettimeofday({1136823753, 233667}, NULL) = 0 select(9, [3], [], [3 8], {0, 0}) = 0 (Timeout) gettimeofday({1136823753, 234002}, NULL) = 0 gettimeofday({1136823753, 234042}, NULL) = 0 gettimeofday({1136823753, 234070}, NULL) = 0 gettimeofday({1136823753, 234110}, NULL) = 0 ioctl(10, FIONBIO, [1]) = 0 gettimeofday({1136823753, 234306}, NULL) = 0 gettimeofday({1136823753, 234330}, NULL) = 0 gettimeofday({1136823753, 234352}, NULL) = 0 gettimeofday({1136823753, 234377}, NULL) = 0 gettimeofday({1136823753, 234410}, NULL) = 0 gettimeofday({1136823753, 234431}, NULL) = 0 gettimeofday({1136823753, 234456}, NULL) = 0 getpeername(10, {sa_family=AF_INET, sin_port=htons(44431), sin_addr=inet_addr("127.0.0.1")}, [16]) = 0 time(NULL) = 1136823753 gettimeofday({1136823753, 234716}, NULL) = 0 gettimeofday({1136823753, 234738}, NULL) = 0 recv(10, "G", 1, MSG_PEEK) = 1 read(10, "GET /3d/?C=M;O=D HTTP/1.1\r\nHost:"..., 2000) = 266 time(NULL) = 1136823753 --- SIGALRM (Alarm clock) @ 0 (0) --- sigreturn() = ? (mask now []) gettimeofday({1136823753, 239059}, NULL) = 0 gettimeofday({1136823753, 239600}, NULL) = 0 gettimeofday({1136823753, 239624}, NULL) = 0 time(NULL) = 1136823753 gettimeofday({1136823753, 243498}, NULL) = 0 select(11, NULL, [10], NULL, {0, 0}) = 1 (out [10], left {0, 0}) write(10, "HTTP/1.1 200 OK\r\nKeep-Alive: tim"..., 190) = 190 gettimeofday({1136823753, 243785}, NULL) = 0 select(11, NULL, [10], NULL, {0, 0}) = 1 (out [10], left {0, 0}) write(10, "<?xml version=\"1.0\" encoding=\"is"..., 3811) = -1 EPIPE (Broken pipe) --- SIGPIPE (Broken pipe) @ 0 (0) --- === immediately followed by writing to file #9 gettimeofday({1136823753, 244020}, NULL) = 0 getpeername(10, 0xbffff9d0, [16]) = -1 ENOTCONN (Transport endpoint is not connected) time(NULL) = 1136823753 setsockopt(10, SOL_SOCKET, SO_LINGER, {onoff=0, linger=0}, 8) = 0 close(10) = 0 write(9, "\10", 1) = 1 write(9, "\0", 1) = 1 write(9, "\0", 1) = 1 write(9, "\0", 1) = 1 write(9, "\1", 1) = 1 write(9, "\t", 1) = 1 _llseek(9, 3579904, [3579904], SEEK_SET) = 0 read(9, "ubclasses.\20\3\37\312\20\5\37\325\4\0\0\0\0\n\0\34\315"..., 1356) = 1356 read(9, "\t", 4096) = 1 _llseek(9, 3581261, [3581261], SEEK_SET) = 0 _llseek(9, 3581261, [3581261], SEEK_SET) = 0 _llseek(9, -1, [3581260], SEEK_CUR) = 0 write(9, "\20", 1) = 1 write(9, "\t", 1) = 1 write(9, "\37", 1) = 1 write(9, "\3", 1) = 1 === 11 seconds later the saving was done (yes storing is that slow): close(9) = 0 munmap(0x40188000, 4096) = 0 gettimeofday({1136823764, 800773}, NULL) = 0 === ... and everything is back to normal ... All this is on Linux, 2.4 kernel, Ian's 3.7.7 VM. Did anybody see something similar? Any ideas? - Bert -

2 2

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

Vm-dev January 2006