I am glad to announce the release of ATS2-0.1.8.
This version is primarily released for the purpose of
building Debian packages. If you have already installed
ATS2-0.1.7, then there is really no need to install this
one.
From this point on, the ‘patsopt’ built based on
the defaut setting will no long invoke GC when called to
compile ATS source. If you need it to run with GC support,
try to build ‘patsopt’ using one of the following command-lines:
./configure && make GCFLAG=-D_ATS_GCBDW # using Boehm-GC
./configure && make GCFLAG=-D_ATS_GCATS # using GCATS (GC for ATS1)
Cheers!
–Hongwei
This is the 17th release of ATS2, the successor of the ATS
programming language. The compiler for ATS2 is given the name
ATS/Positats, ATS2/Postiats or simply Postiats.
Major releases of external packages for ATS2 are available at:
Here is a list of major additions and changes since the last release:
GCATS (GC for ATS1) is no longer used by default to build patsopt.
Some common command-lines for building patsopt are listed below:
./configure && make # using no GC
./configure && make GCFLAG=-D_ATS_GCBDW # using Boehm-GC
./configure && make GCFLAG=-D_ATS_GCATS # using GCATS (GC for ATS1)
New Year. New Location. New Benefits. New Data Center in Ashburn, VA.
GigeNET is offering a free month of service with a new server in Ashburn.
Choose from 2 high performing configs, both with 100TB of bandwidth.
Higher redundancy.Lower latency.Increased capacity.Completely compliant.
That’s not really ATS2 related, rather GCC related. In an older thread,
someone suggested to pass “-fno-aggressive-loop-optimization” in CFLAGS
while building ATS2 with GCC. I did this before with GCC 4.6, but for some
unknown reasons, it does not work any more: GCC 4.6 does not recognize the
option anymore, and I get an error about GCC being unable to produce
binaries during “config”. As I know some people use this option to build
ATS2, I just wanted to tell.Le mardi 27 janvier 2015 17:39:52 UTC+1, gmhwxi a écrit :
Could you tell us how this error can be re-produced?
On Tue, Jan 27, 2015 at 6:26 AM, ‘Yannick Duchêne’ via ats-lang-users < ats-l...@googlegroups.com <javascript:>> wrote:
As a side note, I get an error with a sometime recommended option,
“-ffunction-sections”, which now give me an error. I had to reinstall
everything on my machine after an hard drive failure, but GCC version is
the same as before, that is 4.6 (but I don’t mind, I juste dropped the
option, and just wanted to tell, in case).
Oops, typo, I meant the “-fno-aggressive-loop-optimization” option.
My apologies for the basic question, but I can’t figure any more where to
get ATS2-include. I remember I used to get it, but can’t remember where and
the download page at the ATS2 web site does not tell where.
As a side note, I get an error with a sometime recommended option,
“-ffunction-sections”, which now give me an error. I had to reinstall
everything on my machine after an hard drive failure, but GCC version is
the same as before, that is 4.6 (but I don’t mind, I juste dropped the
option, and just wanted to tell, in case).
I have not tested it, but talking about ATS versions, I was wondering if
a future version will have a “$showsort” just like there already is a
“$showtype”, so that one could do “val _ = $showsort(string)” like one can
already do “val _ = $showtype(“abcd”)”?
As a side note, I get an error with a sometime recommended option,
“-ffunction-sections”, which now give me an error. I had to reinstall
everything on my machine after an hard drive failure, but GCC version is
the same as before, that is 4.6 (but I don’t mind, I juste dropped the
option, and just wanted to tell, in case).
Oops, typo, I meant the “-fno-aggressive-loop-optimization” option.
I don’t know what cause the error, if it’s Contrib 0.1.7 which is not
compatible with Postiats 0.1.9, the STATS file glib/SATS/glib/gmain.sats
in Contrib, does not type‑check, either with patsopt -tc -s or patscc -tcats.
-fno-aggressive-loop-optimizationsOn Tue, Jan 27, 2015 at 1:30 PM, ‘Yannick Duchêne’ via ats-lang-users < ats-lan...@googlegroups.com> wrote:
That’s not really ATS2 related, rather GCC related. In an older thread,
someone suggested to pass “-fno-aggressive-loop-optimization” in CFLAGS
while building ATS2 with GCC. I did this before with GCC 4.6, but for some
unknown reasons, it does not work any more: GCC 4.6 does not recognize the
option anymore, and I get an error about GCC being unable to produce
binaries during “config”. As I know some people use this option to build
ATS2, I just wanted to tell.
Le mardi 27 janvier 2015 17:39:52 UTC+1, gmhwxi a écrit :
Could you tell us how this error can be re-produced?
On Tue, Jan 27, 2015 at 6:26 AM, ‘Yannick Duchêne’ via ats-lang-users < ats-l...@googlegroups.com> wrote:
As a side note, I get an error with a sometime recommended option,
“-ffunction-sections”, which now give me an error. I had to reinstall
everything on my machine after an hard drive failure, but GCC version is
the same as before, that is 4.6 (but I don’t mind, I juste dropped the
option, and just wanted to tell, in case).
Oops, typo, I meant the “-fno-aggressive-loop-optimization” option.
The former. It is about running the compiler (patsopt) with GC. It does not
affect the code generated by the compiler.On Sun, Jan 25, 2015 at 5:26 PM, john skaller <ska...@users.sourceforge.net wrote:
On 26/01/2015, at 4:06 AM, Hongwei Xi wrote:
Hi,
I am glad to announce the release of ATS2-0.1.8.
This version is primarily released for the purpose of
building Debian packages. If you have already installed
ATS2-0.1.7, then there is really no need to install this
one.
From this point on, the ‘patsopt’ built based on
the defaut setting will no long invoke GC when called to
compile ATS source. If you need it to run with GC support,
try to build ‘patsopt’ using one of the following command-lines:
./configure && make GCFLAG=-D_ATS_GCBDW # using Boehm-GC
./configure && make GCFLAG=-D_ATS_GCATS # using GCATS (GC for ATS1)
I’m confused. Is this running the compiler with the GC, or is it
configuring
the program the compiler is generating to run with GC?
New Year. New Location. New Benefits. New Data Center in Ashburn, VA.
GigeNET is offering a free month of service with a new server in Ashburn.
Choose from 2 high performing configs, both with 100TB of bandwidth.
Higher redundancy.Lower latency.Increased capacity.Completely compliant. Best Open Source Mac Front-Ends 2023
My apologies for the basic question, but I can’t figure any more where to
get ATS2-include. I remember I used to get it, but can’t remember where and
the download page at the ATS2 web site does not tell where.
As a side note, I get an error with a sometime recommended option,
“-ffunction-sections”, which now give me an error. I had to reinstall
everything on my machine after an hard drive failure, but GCC version is
the same as before, that is 4.6 (but I don’t mind, I juste dropped the
option, and just wanted to tell, in case).
I don’t know what cause the error, if it’s Contrib 0.1.7 which is not
compatible with Postiats 0.1.9, the STATS file glib/SATS/glib/gmain.sats
in Contrib, does not type‑check, either with patsopt -tc -s or patscc -tcats.
I have not tested it, but talking about ATS versions, I was wondering if a
future version will have a “$showsort” just like there already is a
“$showtype”, so that one could do “val _ = $showsort(string)” like one can
already do “val _ = $showtype(“abcd”)”?
With the hope it can help, I’m attaching an archive containing this:
the DATS file whose compilation terminates in a core dump
a file with the exact command terminating in a core dump
the C file generated before the core dump
one of the funny postiats_funlablst_xxxxxx patscc seems to produce
the config.log file from ATS’s build
a file with GCC version informations
a file with CFLAGS informationsLe vendredi 30 janvier 2015 11:39:00 UTC+1, Yannick Duchêne a écrit :
If it’s added to CFLAGS before configure, then configure fails. If it’s
added after configure and before make, it does not fail, but I’m afraid its
not applied.
GCC version is: gcc (Ubuntu/Linaro 4.6.3-1ubuntu5) 4.6.3.
I don’t know why the option make configure fails and do not with make.
Whatever, patscc crashes. Will see later if it’s random or not (the last
time I encountered the same, it was random).
Le vendredi 30 janvier 2015 11:16:24 UTC+1, Yannick Duchêne a écrit :
I get a core dump again, even when ATS is compiled with this option. And
I’m getting it with a very simple tutorial file.
Le mardi 27 janvier 2015 19:36:29 UTC+1, gmhwxi a écrit :
If it’s added to CFLAGS before configure, then configure fails. If it’s
added after configure and before make, it does not fail, but I’m afraid its
not applied.
GCC version is: gcc (Ubuntu/Linaro 4.6.3-1ubuntu5) 4.6.3.
I don’t know why the option make configure fails and do not with make.
Whatever, patscc crashes. Will see later if it’s random or not (the last
time I encountered the same, it was random).Le vendredi 30 janvier 2015 11:16:24 UTC+1, Yannick Duchêne a écrit :
I get a core dump again, even when ATS is compiled with this option. And
I’m getting it with a very simple tutorial file.
Le mardi 27 janvier 2015 19:36:29 UTC+1, gmhwxi a écrit :
I have added a README file that contains the note.On Tuesday, February 17, 2015 at 6:38:48 PM UTC-5, Yannick Duchêne wrote:
Le mercredi 18 février 2015 00:12:21 UTC+1, gmhwxi a écrit :
These file are not meant to be used independently.
They are included in the file ./…/glib.sats
I have added a note to clarify this issue.
A note in a README in this contrib’s directory?
I have issue with near to everything in contrib/GTK. I checked $PATSHOMERELOC is OK, and it is (points to ATS installation directory).
Is this something similar or related?
These file are not meant to be used independently.
They are included in the file ./…/glib.sats
I have added a note to clarify this issue.
A note in a README in this contrib’s directory?
I have issue with near to everything in contrib/GTK. I checked $PATSHOMERELOC is OK, and it is (points to ATS installation directory).
Is this something similar or related?