pharo-users@lists.pharo.org

Any question about pharo is welcome

View all threads

Pharo VM Release - v9.0.21

T
tesonep@gmail.com
Mon, Dec 12, 2022 1:30 PM

Hello,
I have released a new version of the Pharo VM for Pharo 9, Pharo 10 and
Pharo 11. This VM is accessible right now from Zero-Conf, updating it in
the Pharo Launcher or using the usual downloads (as described in
pharo.org/download).

This version includes a series of bug fixes and upgrades on the third-party
libraries.

Changelog:

  • Implementing High resolution clock for ARM64 (Used during profiling)
  • Updating third party libraries for all the graphic layer.
  • Fixing a performance regression on the allocation of opcodes and fix-ups.
    Cleaning only the ones that are going to be used.
    Like this, this version has the same speed than before when
    allocating in the stack.
  • Correctly handling the encoding of the command line arguments of the VM
    (Windows)
  • Allocating the opcodes and fixup structs only once and reusing them
    (Reducing risk of C Stack Overflow)

Thanks a lot, and any doubt please let me know.
Cheers,

Pablo on behalf of the whole Pharo team.

--
Pablo Tesone.
tesonep@gmail.com

Hello, I have released a new version of the Pharo VM for Pharo 9, Pharo 10 and Pharo 11. This VM is accessible right now from Zero-Conf, updating it in the Pharo Launcher or using the usual downloads (as described in pharo.org/download). This version includes a series of bug fixes and upgrades on the third-party libraries. Changelog: - Implementing High resolution clock for ARM64 (Used during profiling) - Updating third party libraries for all the graphic layer. - Fixing a performance regression on the allocation of opcodes and fix-ups. Cleaning only the ones that are going to be used. Like this, this version has the same speed than before when allocating in the stack. - Correctly handling the encoding of the command line arguments of the VM (Windows) - Allocating the opcodes and fixup structs only once and reusing them (Reducing risk of C Stack Overflow) Thanks a lot, and any doubt please let me know. Cheers, Pablo on behalf of the whole Pharo team. -- Pablo Tesone. tesonep@gmail.com
GP
Guillermo Polito
Mon, Dec 12, 2022 1:31 PM

Thanks Pablo!!!

El 12 dic. 2022, a las 14:30, tesonep@gmail.com escribió:

Hello,
I have released a new version of the Pharo VM for Pharo 9, Pharo 10 and Pharo 11. This VM is accessible right now from Zero-Conf, updating it in the Pharo Launcher or using the usual downloads (as described in pharo.org/download http://pharo.org/download).

This version includes a series of bug fixes and upgrades on the third-party libraries.

Changelog:

  • Implementing High resolution clock for ARM64 (Used during profiling)
  • Updating third party libraries for all the graphic layer.
  • Fixing a performance regression on the allocation of opcodes and fix-ups.
    Cleaning only the ones that are going to be used.
    Like this, this version has the same speed than before when allocating in the stack.
  • Correctly handling the encoding of the command line arguments of the VM (Windows)
  • Allocating the opcodes and fixup structs only once and reusing them (Reducing risk of C Stack Overflow)

Thanks a lot, and any doubt please let me know.
Cheers,

Pablo on behalf of the whole Pharo team.

--
Pablo Tesone.
tesonep@gmail.com mailto:tesonep@gmail.com

Thanks Pablo!!! > El 12 dic. 2022, a las 14:30, tesonep@gmail.com escribió: > > Hello, > I have released a new version of the Pharo VM for Pharo 9, Pharo 10 and Pharo 11. This VM is accessible right now from Zero-Conf, updating it in the Pharo Launcher or using the usual downloads (as described in pharo.org/download <http://pharo.org/download>). > > This version includes a series of bug fixes and upgrades on the third-party libraries. > > Changelog: > > - Implementing High resolution clock for ARM64 (Used during profiling) > - Updating third party libraries for all the graphic layer. > - Fixing a performance regression on the allocation of opcodes and fix-ups. > Cleaning only the ones that are going to be used. > Like this, this version has the same speed than before when allocating in the stack. > - Correctly handling the encoding of the command line arguments of the VM (Windows) > - Allocating the opcodes and fixup structs only once and reusing them (Reducing risk of C Stack Overflow) > > Thanks a lot, and any doubt please let me know. > Cheers, > > Pablo on behalf of the whole Pharo team. > > -- > Pablo Tesone. > tesonep@gmail.com <mailto:tesonep@gmail.com>
NH
Norbert Hartl
Mon, Dec 12, 2022 4:24 PM

Thanks Pablo for the new vm but as well for the update notification. This is super valuable.

Norbert

Am 12.12.2022 um 14:30 schrieb tesonep@gmail.com:

Hello,
I have released a new version of the Pharo VM for Pharo 9, Pharo 10 and Pharo 11. This VM is accessible right now from Zero-Conf, updating it in the Pharo Launcher or using the usual downloads (as described in pharo.org/download http://pharo.org/download).

This version includes a series of bug fixes and upgrades on the third-party libraries.

Changelog:

  • Implementing High resolution clock for ARM64 (Used during profiling)
  • Updating third party libraries for all the graphic layer.
  • Fixing a performance regression on the allocation of opcodes and fix-ups.
    Cleaning only the ones that are going to be used.
    Like this, this version has the same speed than before when allocating in the stack.
  • Correctly handling the encoding of the command line arguments of the VM (Windows)
  • Allocating the opcodes and fixup structs only once and reusing them (Reducing risk of C Stack Overflow)

Thanks a lot, and any doubt please let me know.
Cheers,

Pablo on behalf of the whole Pharo team.

--
Pablo Tesone.
tesonep@gmail.com mailto:tesonep@gmail.com

Thanks Pablo for the new vm but as well for the update notification. This is super valuable. Norbert > Am 12.12.2022 um 14:30 schrieb tesonep@gmail.com: > > Hello, > I have released a new version of the Pharo VM for Pharo 9, Pharo 10 and Pharo 11. This VM is accessible right now from Zero-Conf, updating it in the Pharo Launcher or using the usual downloads (as described in pharo.org/download <http://pharo.org/download>). > > This version includes a series of bug fixes and upgrades on the third-party libraries. > > Changelog: > > - Implementing High resolution clock for ARM64 (Used during profiling) > - Updating third party libraries for all the graphic layer. > - Fixing a performance regression on the allocation of opcodes and fix-ups. > Cleaning only the ones that are going to be used. > Like this, this version has the same speed than before when allocating in the stack. > - Correctly handling the encoding of the command line arguments of the VM (Windows) > - Allocating the opcodes and fixup structs only once and reusing them (Reducing risk of C Stack Overflow) > > Thanks a lot, and any doubt please let me know. > Cheers, > > Pablo on behalf of the whole Pharo team. > > -- > Pablo Tesone. > tesonep@gmail.com <mailto:tesonep@gmail.com>
SM
Stewart MacLean
Wed, Dec 14, 2022 10:08 PM

Hi Pablo,

I'm on Pharo 10 and intermittently experience "the spinning wheel" when
making LibC calls. The spin dump is below.

I'm wondering if the last fix:

  • Allocating the opcodes and fixup structs only once and reusing them
    (Reducing risk of C Stack Overflow))

would potentially rectify this?

Thanks to you and the whole team for all the great work on the machine!

Stewart

Thread 0x9a54d    1000 samples (1-1000)    priority 46 (base 46)
1000  ??? [0x1062e1a78]
1000  ??? [0x1063c2374]
1000  ??? [0x1062e1ab8]
1000  ??? [0x10637a0e0]
1000  ??? [0x10637a324]
1000  ??? [0x1062e27f4]
1000  ??? [0x10637763c]
1000  ??? [0x1062e070c]
1000  ceStackOverflow + 196 (libPharoVMCore.dylib +
451716) [0x10453a484]
1000  checkForEventsMayContextSwitch + 828
(libPharoVMCore.dylib + 430696) [0x104535268]
1000  ioProcessEvents + 16 (libPharoVMCore.dylib +
25800) [0x1044d24c8]
1000  aioPoll + 80 (libPharoVMCore.dylib + 788412)
[0x10458c7bc]
1000  _sigtramp + 56 (libsystem_platform.dylib +
17636) [0x1a20a84e4]
1000  handleSignal + 80
(libUnixOSProcessPlugin.dylib + 46264) [0x11cf674b8]
1000  signalSemaphoreWithIndex + 264
(libPharoVMCore.dylib + 786524) [0x10458c05c]
1000  aioInterruptPoll + 104
(libPharoVMCore.dylib + 789424) [0x10458cbb0]
1000  platform_semaphore_wait + 20
(libPharoVMCore.dylib + 36804) [0x1044d4fc4]
1000  _dispatch_semaphore_wait_slow +
132 (libdispatch.dylib + 18420) [0x1a1ed37f4]
1000  <patched truncated backtrace>
1000  semaphore_wait_trap + 8
(libsystem_kernel.dylib + 6544) [0x1a2055990]
*1000  ??? (kernel.release.t8101 +
575116) [0xfffffe000799868c]

On Tue, Dec 13, 2022 at 2:30 AM tesonep@gmail.com tesonep@gmail.com wrote:

Hello,
I have released a new version of the Pharo VM for Pharo 9, Pharo 10 and
Pharo 11. This VM is accessible right now from Zero-Conf, updating it in
the Pharo Launcher or using the usual downloads (as described in
pharo.org/download).

This version includes a series of bug fixes and upgrades on the
third-party libraries.

Changelog:

  • Implementing High resolution clock for ARM64 (Used during profiling)
  • Updating third party libraries for all the graphic layer.
  • Fixing a performance regression on the allocation of opcodes and fix-ups.
    Cleaning only the ones that are going to be used.
    Like this, this version has the same speed than before when
    allocating in the stack.
  • Correctly handling the encoding of the command line arguments of the VM
    (Windows)
  • Allocating the opcodes and fixup structs only once and reusing them
    (Reducing risk of C Stack Overflow)

Thanks a lot, and any doubt please let me know.
Cheers,

Pablo on behalf of the whole Pharo team.

--
Pablo Tesone.
tesonep@gmail.com

Hi Pablo, I'm on Pharo 10 and intermittently experience "the spinning wheel" when making LibC calls. The spin dump is below. I'm wondering if the last fix: - Allocating the opcodes and fixup structs only once and reusing them (Reducing risk of C Stack Overflow)) would potentially rectify this? Thanks to you and the whole team for all the great work on the machine! Stewart Thread 0x9a54d 1000 samples (1-1000) priority 46 (base 46) 1000 ??? [0x1062e1a78] 1000 ??? [0x1063c2374] 1000 ??? [0x1062e1ab8] 1000 ??? [0x10637a0e0] 1000 ??? [0x10637a324] 1000 ??? [0x1062e27f4] 1000 ??? [0x10637763c] 1000 ??? [0x1062e070c] 1000 ceStackOverflow + 196 (libPharoVMCore.dylib + 451716) [0x10453a484] 1000 checkForEventsMayContextSwitch + 828 (libPharoVMCore.dylib + 430696) [0x104535268] 1000 ioProcessEvents + 16 (libPharoVMCore.dylib + 25800) [0x1044d24c8] 1000 aioPoll + 80 (libPharoVMCore.dylib + 788412) [0x10458c7bc] 1000 _sigtramp + 56 (libsystem_platform.dylib + 17636) [0x1a20a84e4] 1000 handleSignal + 80 (libUnixOSProcessPlugin.dylib + 46264) [0x11cf674b8] 1000 signalSemaphoreWithIndex + 264 (libPharoVMCore.dylib + 786524) [0x10458c05c] 1000 aioInterruptPoll + 104 (libPharoVMCore.dylib + 789424) [0x10458cbb0] 1000 platform_semaphore_wait + 20 (libPharoVMCore.dylib + 36804) [0x1044d4fc4] 1000 _dispatch_semaphore_wait_slow + 132 (libdispatch.dylib + 18420) [0x1a1ed37f4] 1000 <patched truncated backtrace> 1000 semaphore_wait_trap + 8 (libsystem_kernel.dylib + 6544) [0x1a2055990] *1000 ??? (kernel.release.t8101 + 575116) [0xfffffe000799868c] On Tue, Dec 13, 2022 at 2:30 AM tesonep@gmail.com <tesonep@gmail.com> wrote: > Hello, > I have released a new version of the Pharo VM for Pharo 9, Pharo 10 and > Pharo 11. This VM is accessible right now from Zero-Conf, updating it in > the Pharo Launcher or using the usual downloads (as described in > pharo.org/download). > > This version includes a series of bug fixes and upgrades on the > third-party libraries. > > Changelog: > > - Implementing High resolution clock for ARM64 (Used during profiling) > - Updating third party libraries for all the graphic layer. > - Fixing a performance regression on the allocation of opcodes and fix-ups. > Cleaning only the ones that are going to be used. > Like this, this version has the same speed than before when > allocating in the stack. > - Correctly handling the encoding of the command line arguments of the VM > (Windows) > - Allocating the opcodes and fixup structs only once and reusing them > (Reducing risk of C Stack Overflow) > > Thanks a lot, and any doubt please let me know. > Cheers, > > Pablo on behalf of the whole Pharo team. > > -- > Pablo Tesone. > tesonep@gmail.com >