xst fails...

Hi,

I'm currently trying to synthesize a big design on a Virtex4-VLX100. Now
the problem is, that xst fails and just gives the following line:

Process "Synthesize" failed

Is there a way to hunt for the problem that causes this behavior? I'm a
little bit lost, since there is no hint by xst.

Thanks,
Matthias

On 14 Aug, 10:31, Matthias Alles <REMOVEallesCAPIT...@NOeit.SPAMuni-
kl.de> wrote:
> Hi,
>
> I'm currently trying to synthesize a big design on a Virtex4-VLX100. Now
> the problem is, that xst fails and just gives the following line:
>
> Process "Synthesize" failed
>
> Is there a way to hunt for the problem that causes this behavior? I'm a
> little bit lost, since there is no hint by xst.
>
> Thanks,
> Matthias

Try posting the full output.

Cheers,
Jon

On Tue, 14 Aug 2007 02:44:20 -0700, Jon Beniston <jon@beniston.com> wrote:

>> I'm currently trying to synthesize a big design on a Virtex4-VLX100. Now
>> the problem is, that xst fails and just gives the following line:
>>
>> Process "Synthesize" failed
>>
>> Is there a way to hunt for the problem that causes this behavior? I'm a
>> little bit lost, since there is no hint by xst.

>Try posting the full output.


His problem is: this is the full output.

I had cases that were even less verbose: ""


Sometimes it helps to do project->cleanup project files, perhaps with touching
the sources to update the date.

Sometimes I had to rebuilt the project in a different directory under a different name,
source file by source file. That was not funny.

One extra tough problem simply vanished into thin air when I updated 
from ISE 8.2 to 9.1

Last week I changed some VHDL components to direct entity instantiations and 
made a typo in an entity name. I got no error during compilation but a
crash when linking/building the netlist.

So, it could be about everyting.

It would be a great help if ISE could spit out a makefile.


regards, Gerhard

There is nothing of interest in the output:

INFO:Xst:2261 - The FF/Latch <configure.parallelism_2> in Unit <hws> is
equivalent to the following 155 FFs/Latches, which will be removed :
<configure.parallelism_3> <configure.parallelism_4>
<configure.parallelism_5> <configure.use_virtual_edge>
<configure.cnd_123> <configure.cnd_122> <configure.cnd_121>
<configure.cnd_120> <configure.cnd_119> <configure.cnd_118>
<configure.cnd_117> <configure.cnd_116> <configure.cnd_115>
<configure.cnd_114> <configure.cnd_113> <configure.cnd_112> <configure.cn
d_111> <configure.cnd_110> <configure.cnd_109> <configure.cnd_108>
<configure.cnd_107> <configure.cnd_106> <configure.cnd_105>
<configure.cnd_104> <configure.cnd_103> <configure.cnd_102>
<configure.cnd_101> <configure.cnd_100> <configure.cnd_99>
<configure.cnd_98> <configure.cnd_97> <configure.cnd_96>
<configure.cnd_95> <configure.cnd_94> <configure.cnd_93>
<configure.cnd_92> <configure.cnd_91> <configure.cnd_90>
<configure.cnd_89> <configure.cnd_88> <configure.cnd_87>
<configure.cnd_86> <configu
re.cnd_85>
   <configure.cnd_84> <configure.cnd_83> <configure.cnd_82>
<configure.cnd_81> <configure.cnd_80> <configure.cnd_79>
<configure.cnd_78> <configure.cnd_77> <configure.cnd_76>
<configure.cnd_75> <configure.cnd_74> <configure.cnd_73>
<configure.cnd_72> <configure.cnd_71> <configure.cnd_70>
<configure.cnd_69> <configure.cnd_68> <configure.cnd_67>
<configure.cnd_66> <configure.cnd_65> <configure.cnd_64>
<configure.cnd_63> <configure.cnd_62> <configure.cnd_61>
<configure.cnd_60> <configure.cnd_59> <con
figure.cnd_58> <configure.cnd_57> <configure.cnd_56> <configure.cnd_55>
<configure.cnd_54> <configure.cnd_53> <configure.cnd_52>
<configure.cnd_51> <configure.cnd_50> <configure.cnd_49>
<configure.cnd_48> <configure.cnd_47> <configure.cnd_46>
<configure.cnd_45> <configure.cnd_44> <configure.cnd_43>
<configure.cnd_42> <configure.cnd_41> <configure.cnd_40>
<configure.cnd_39> <configure.cnd_38> <configure.cnd_37>
<configure.cnd_36> <configure.cnd_35> <configure.cnd_34>
<configure.cnd_33> <configure.
cnd_32>
   <configure.cnd_31> <configure.cnd_30> <configure.cnd_29>
<configure.cnd_28> <configure.cnd_27> <configure.cnd_26>
<configure.cnd_25> <configure.cnd_24> <configure.cnd_23>
<configure.cnd_22> <configure.cnd_21> <configure.cnd_20>
<configure.cnd_19> <configure.cnd_18> <configure.cnd_17>
<configure.cnd_16> <configure.cnd_15> <configure.cnd_14>
<configure.cnd_13> <configure.cnd_12> <configure.cnd_11>
<configure.cnd_10> <configure.cnd_9> <configure.cnd_8> <configure.cnd_7>
<configure.cnd_6> <configu
re.cnd_5> <configure.cnd_4> <configure.cnd_3> <configure.cnd_2>
<configure.cnd_1> <configure.cnd_0> <configure.cnd_high_1>
<configure.cnd_high_2> <configure.cnd_high_3> <configure.cnd_high_4>
<configure.nr_msg_per_node_3> <configure.nr_msg_per_node_5>
<configure.nr_msg_per_node_7> <configure.nr_msg_per_node_10>
<configure.nr_msg_per_node_11> <configure.nr_chv_per_node_3>
<configure.nr_chv_per_node_4> <configure.nr_chv_per_node_5>
<configure.nr_chv_per_node_10> <configure.virtual_edge_pos_0>
   <configure.virtual_edge_pos_2> <configure.virtual_edge_pos_3>
<configure.virtual_edge_pos_4> <configure.mean_bound_VNR_10>
<configure.mean_bound_VNR_13> <configure.max_iteration_0>
<configure.max_iteration_1> <configure.max_iteration_2>
<configure.max_iteration_3> <configure.cns_per_node_2>
<configure.cns_per_node_3> <configure.cns_per_node_5>
<configure.cns_per_node_6>

Process "Synthesize" failed

That's all! The info is correct, since the values are constant.


Jon Beniston schrieb:
> On 14 Aug, 10:31, Matthias Alles <REMOVEallesCAPIT...@NOeit.SPAMuni-
> kl.de> wrote:
>> Hi,
>>
>> I'm currently trying to synthesize a big design on a Virtex4-VLX100. Now
>> the problem is, that xst fails and just gives the following line:
>>
>> Process "Synthesize" failed
>>
>> Is there a way to hunt for the problem that causes this behavior? I'm a
>> little bit lost, since there is no hint by xst.
>>
>> Thanks,
>> Matthias
> 
> Try posting the full output.
> 
> Cheers,
> Jon
> 

At least updating to the newest version didn't help. I'm currently using
ISE 9.2.01i. I will try with the cleaned project and maybe start a new
project as well.

Matthias

Gerhard Hoffmann schrieb:
> On Tue, 14 Aug 2007 02:44:20 -0700, Jon Beniston <jon@beniston.com> wrote:
> 
>>> I'm currently trying to synthesize a big design on a Virtex4-VLX100. Now
>>> the problem is, that xst fails and just gives the following line:
>>>
>>> Process "Synthesize" failed
>>>
>>> Is there a way to hunt for the problem that causes this behavior? I'm a
>>> little bit lost, since there is no hint by xst.
> 
>> Try posting the full output.
> 
> 
> His problem is: this is the full output.
> 
> I had cases that were even less verbose: ""
> 
> 
> Sometimes it helps to do project->cleanup project files, perhaps with touching
> the sources to update the date.
> 
> Sometimes I had to rebuilt the project in a different directory under a different name,
> source file by source file. That was not funny.
> 
> One extra tough problem simply vanished into thin air when I updated 
> from ISE 8.2 to 9.1
> 
> Last week I changed some VHDL components to direct entity instantiations and 
> made a typo in an entity name. I got no error during compilation but a
> crash when linking/building the netlist.
> 
> So, it could be about everyting.
> 
> It would be a great help if ISE could spit out a makefile.
> 
> 
> regards, Gerhard
> 

On Tue, 14 Aug 2007 12:32:30 +0200, Matthias Alles
<REMOVEallesCAPITALS@NOeit.SPAMuni-kl.de> wrote:

>There is nothing of interest in the output:
>
>INFO:Xst:2261 - The FF/Latch <configure.parallelism_2> in Unit <hws> is
>equivalent to the following 155 FFs/Latches, which will be removed :


Anything in the synthesis report file (.syr)?

Since the error appears to be near register removal, it  may be worth
turning OFF "equivalent register removal" and seeing if the error
remains.

- Brian

I tried this, however this still doesn't work.
I started xst now from the console and now I see that I get a memory
violation at that certain point.

I will use Precision from Mentor Graphics from now on for the synthesis.

Matthias


Brian Drummond schrieb:
> On Tue, 14 Aug 2007 12:32:30 +0200, Matthias Alles
> <REMOVEallesCAPITALS@NOeit.SPAMuni-kl.de> wrote:
> 
>> There is nothing of interest in the output:
>>
>> INFO:Xst:2261 - The FF/Latch <configure.parallelism_2> in Unit <hws> is
>> equivalent to the following 155 FFs/Latches, which will be removed :
> 
> 
> Anything in the synthesis report file (.syr)?
> 
> Since the error appears to be near register removal, it  may be worth
> turning OFF "equivalent register removal" and seeing if the error
> remains.
> 
> - Brian