... at http://www.mirandabanda.org/files/Cog/VM/VM.r3143.
CogVM binaries as per VMMaker.oscog-tpr.949/r3143
Spur: Fix regression in primitiveNewWithArg from VMMaker.oscog-eem.859 which caused basicNew: on fixed classes to raise OutOfMemory instead of the relevant error.
Have checkForAndFollowForwardedPrimitiveState follow other references in the frame if a forwarder is found on the stack.
All: Add a first cut at the primitiveSlotAt[Put] primitives.
Reimplement primitiveBitShift for 64-bits.
Define areIntegers:and: for 64-bit Spur, and move it into the ObjectMemory hierarchies. Add maxSmallInteger and minSmallInteger and matching manifest constants. Use these in LargeIntegersPlugin and Matrix2x3Plugin.
Use numBytesOf: instead of lengthOf: in the 64-bit integer conversion routines. Reorder cases in Spur's numBytesOf: to put common case first.
The linux versions (ht and normal) seem to be broken with the latest Trunk image[1] (and any other image I tried). I suspect that the rendering code is causing this, because the image becomes unresponsive right after some UI changes, and uses 100% CPU. I tried to open a browser, but got a few red squares with yellow crosses along with some blurred gray lines instead. Sometimes it stops responding after the first mouse click.
Levente
[1] http://build.squeak.org/job/SqueakTrunk/lastBuild/artifact/target/TrunkImage...
On Thu, 20 Nov 2014, Eliot Miranda wrote:
... at http://www.mirandabanda.org/files/Cog/VM/VM.r3143. CogVM binaries as per VMMaker.oscog-tpr.949/r3143
Spur: Fix regression in primitiveNewWithArg from VMMaker.oscog-eem.859 which caused basicNew: on fixed classes to raise OutOfMemory instead of the relevant error.
Have checkForAndFollowForwardedPrimitiveState follow other references in the frame if a forwarder is found on the stack.
All: Add a first cut at the primitiveSlotAt[Put] primitives.
Reimplement primitiveBitShift for 64-bits.
Define areIntegers:and: for 64-bit Spur, and move it into the ObjectMemory hierarchies. Add maxSmallInteger and minSmallInteger and matching manifest constants. Use these in LargeIntegersPlugin and Matrix2x3Plugin.
Use numBytesOf: instead of lengthOf: in the 64-bit integer conversion routines. Reorder cases in Spur's numBytesOf: to put common case first. -- best,Eliot
On Fri, Nov 21, 2014 at 9:18 AM, Levente Uzonyi leves@elte.hu wrote:
The linux versions (ht and normal) seem to be broken with the latest Trunk image[1] (and any other image I tried). I suspect that the rendering code is causing this, because the image becomes unresponsive right after some UI changes, and uses 100% CPU. I tried to open a browser, but got a few red squares with yellow crosses along with some blurred gray lines instead. Sometimes it stops responding after the first mouse click.
Yes I can reproduce this. Investigating now.
Levente
[1] http://build.squeak.org/job/SqueakTrunk/lastBuild/ artifact/target/TrunkImage.zip
On Thu, 20 Nov 2014, Eliot Miranda wrote:
... at http://www.mirandabanda.org/files/Cog/VM/VM.r3143.
CogVM binaries as per VMMaker.oscog-tpr.949/r3143
Spur: Fix regression in primitiveNewWithArg from VMMaker.oscog-eem.859 which caused basicNew: on fixed classes to raise OutOfMemory instead of the relevant error.
Have checkForAndFollowForwardedPrimitiveState follow other references in the frame if a forwarder is found on the stack.
All: Add a first cut at the primitiveSlotAt[Put] primitives.
Reimplement primitiveBitShift for 64-bits.
Define areIntegers:and: for 64-bit Spur, and move it into the ObjectMemory hierarchies. Add maxSmallInteger and minSmallInteger and matching manifest constants. Use these in LargeIntegersPlugin and Matrix2x3Plugin.
Use numBytesOf: instead of lengthOf: in the 64-bit integer conversion routines. Reorder cases in Spur's numBytesOf: to put common case first. -- best,Eliot
This is a touch puzzling. I’m trying to work out how the changes to BItBltPlugin I checked in last night could possibly cause this and I’m coming up blank.
The actual changes were - add an extra bit of init to the initialiseModule code for builds where ENABLE_FAST_BLT is true. This code was already in the plain interp code as of about a year ago and somehow got missed in the cog code. add a new prim stub, again mostly wrapped in ENABLE_FAST_BLT so that it should be null for most platforms.
Also update Ben’s fast blt for ARM code, which has support for the new prim.
I could imagine ARM vms having problems if something went wrong - maybe I managed to miss a file? - but other platforms? That’s nuts.
tim -- tim Rowledge; tim@rowledge.org; http://www.rowledge.org/tim "Bother" said Pooh, when Piglet stubbed his fag out on the semtex.
On Fri, Nov 21, 2014 at 11:06 AM, tim Rowledge tim@rowledge.org wrote:
This is a touch puzzling. I’m trying to work out how the changes to BItBltPlugin I checked in last night could possibly cause this and I’m coming up blank.
I broke bitShift:. Give me a while...
The actual changes were - add an extra bit of init to the initialiseModule code for builds where ENABLE_FAST_BLT is true. This code was already in the plain interp code as of about a year ago and somehow got missed in the cog code. add a new prim stub, again mostly wrapped in ENABLE_FAST_BLT so that it should be null for most platforms.
Also update Ben’s fast blt for ARM code, which has support for the new prim.
I could imagine ARM vms having problems if something went wrong - maybe I managed to miss a file? - but other platforms? That’s nuts.
tim
tim Rowledge; tim@rowledge.org; http://www.rowledge.org/tim "Bother" said Pooh, when Piglet stubbed his fag out on the semtex.
On Fri, Nov 21, 2014 at 9:18 AM, Levente Uzonyi leves@elte.hu wrote:
The linux versions (ht and normal) seem to be broken with the latest Trunk image[1] (and any other image I tried). I suspect that the rendering code is causing this, because the image becomes unresponsive right after some UI changes, and uses 100% CPU. I tried to open a browser, but got a few red squares with yellow crosses along with some blurred gray lines instead. Sometimes it stops responding after the first mouse click.
Yes, it appears I broke the bitSHift: primitive when I rewrite it. Give me a few hours and things should be back to normal...
Levente
[1] http://build.squeak.org/job/SqueakTrunk/lastBuild/ artifact/target/TrunkImage.zip
On Thu, 20 Nov 2014, Eliot Miranda wrote:
... at http://www.mirandabanda.org/files/Cog/VM/VM.r3143.
CogVM binaries as per VMMaker.oscog-tpr.949/r3143
Spur: Fix regression in primitiveNewWithArg from VMMaker.oscog-eem.859 which caused basicNew: on fixed classes to raise OutOfMemory instead of the relevant error.
Have checkForAndFollowForwardedPrimitiveState follow other references in the frame if a forwarder is found on the stack.
All: Add a first cut at the primitiveSlotAt[Put] primitives.
Reimplement primitiveBitShift for 64-bits.
Define areIntegers:and: for 64-bit Spur, and move it into the ObjectMemory hierarchies. Add maxSmallInteger and minSmallInteger and matching manifest constants. Use these in LargeIntegersPlugin and Matrix2x3Plugin.
Use numBytesOf: instead of lengthOf: in the 64-bit integer conversion routines. Reorder cases in Spur's numBytesOf: to put common case first. -- best,Eliot
squeak-dev@lists.squeakfoundation.org