Re: Rootcint's include paths

From: Alexander Mann <amann_at_uni-goettingen.de>
Date: Sat, 25 Sep 2010 16:05:16 +0200

Hi Axel,

yes, ROOTSYS is set and $ROOTSYS/include points to the correct directory, see

http://voatlas20.cern.ch:25980/monitor/logs/4722a647-cdcc-4398-9d3a-79880b6e21b8/tarball_PandaJob_1118401439_ANALY_GRIF-LPNHE/prun_stdout.txt

(Search für "ROOTSYS=" and "No such".)

As I cannot check for the cause of the error directly on the GRID site, I guess I should just always pass -I$ROOTSYS/include to rootcint (just to make sure).

cu,
Alexander

Am 25.09.2010 15:02, schrieb Axel Naumann:
> Hi Alexander,
>
> rootcint has an implicit -I$ROOTSYS/include (or whatever was specified
> as include dir to configure, when configuring with --prefix). Is ROOTSYS
> set, and does $ROOTSYS/include point to the correct directory?
>
> Cheers, Axel.
>
> On 9/25/10 1:58 PM, Alexander Mann wrote:
>>
>> Dear ROOT experts,
>>
>> where does rootcint get the path from, under which it looks for the
>> following files:
>>
>> {TObject.h,TMemberInspector.h,TSelector.h,TString.h,Rtypes.h,TList.h,TNamed.h,TError.h}
>>
>>
>> Motivation for my question:
>>
>> I have a GRID job which has two build jobs on two different sites, and
>> one of them fails whereas the other doesn't. The one which fails says
>>
>> Generating dictionary src/SFrameCore_Dict.cxx
>> Compiling app/sframe_main.cxx
>> In file included from /tmp/weSVSq_cint.cxx:1:
>> /tmp/rootcint_v1FyZB.h:1:21: error: TObject.h: No such file or directory
>> /tmp/rootcint_v1FyZB.h:2:30: error: TMemberInspector.h: No such file or
>> directory
>> In file included from ./include/ISCycleBase.h:18,
>> from /tmp/rootcint_v1FyZB.h:3,
>> from /tmp/weSVSq_cint.cxx:1:
>> ./include/SCycleBaseExec.h:21:23: error: TSelector.h: No such file or
>> directory
>> ./include/SCycleBaseExec.h:22:21: error: TString.h: No such file or
>> directory
>> In file included from ./include/SCycleBaseExec.h:27,
>> from ./include/ISCycleBase.h:18,
>> from /tmp/rootcint_v1FyZB.h:3,
>> from /tmp/weSVSq_cint.cxx:1:
>> ./include/ISCycleBaseNTuple.h:21:20: error: Rtypes.h: No such file or
>> directory
>> In file included from ./include/SCycleBase.h:21,
>> from /tmp/rootcint_v1FyZB.h:8,
>> from /tmp/weSVSq_cint.cxx:1:
>> ./include/SCycleBaseConfig.h:23:19: error: TList.h: No such file or
>> directory
>> In file included from ./include/SCycleBaseConfig.h:29,
>> from ./include/SCycleBase.h:21,
>> from /tmp/rootcint_v1FyZB.h:8,
>> from /tmp/weSVSq_cint.cxx:1:
>> ./include/SCycleConfig.h:23:20: error: TNamed.h: No such file or directory
>> In file included from /tmp/rootcint_v1FyZB.h:19,
>> from /tmp/weSVSq_cint.cxx:1:
>> ./include/SErrorHandler.h:18:20: error: TError.h: No such file or directory
>> Error: external preprocessing failed. :0:
>> !!!Removing src/SFrameCore_Dict.cxx src/SFrameCore_Dict.h !!!
>> Error:
>> /gpfs/automountdir/lpnhe/software/atlas/prod/releases/rel_15-23/DetCommon/15.8.0/InstallArea/i686-slc5-gcc43-opt/bin/../../../../../sw/lcg/app/releases/ROOT/5.22.00j/i686-slc5-gcc43-opt/root/bin/rootcint:
>> error loading headers...
>>
>> On the other hand according to the site admin the files are there:
>>
>>> $ cd
>> /gpfs/automountdir/lpnhe/software/atlas/prod/releases/rel_15-23/DetCommon/15.8.0/InstallArea/i686-slc5-gcc43-opt/bin/../../../../../sw/lcg/app/releases/ROOT/5.22.00j/i686-slc5-gcc43-opt/root/include/
>>
>>> $ ls -l
>> {TObject.h,TMemberInspector.h,TSelector.h,TString.h,Rtypes.h,TList.h,TNamed.h,TError.h}
>>
>>> -rw-r--r-- 1 atls atlas 16290 May 3 18:56 Rtypes.h
>>> -rw-r--r-- 1 atls atlas 3290 May 3 18:56 TError.h
>>> -rw-r--r-- 1 atls atlas 8905 May 3 18:56 TList.h
>>> -rw-r--r-- 1 atls atlas 1770 May 3 18:56 TMemberInspector.h
>>> -rw-r--r-- 1 atls atlas 2642 May 3 18:56 TNamed.h
>>> -rw-r--r-- 1 atls atlas 9303 May 3 18:56 TObject.h
>>> -rw-r--r-- 1 atls atlas 3476 May 3 18:56 TSelector.h
>>> -rw-r--r-- 1 atls atlas 29028 May 3 18:56 TString.h
>>> Related issue :
>>> CC recipient :
>>
>> So why doesn't rootcint find them? It is not passed any include
>> directories, so it does have to take the files it's looking for (or
>> which a subsequent process is looking for) from some other environment
>> variable.
>>
>> cu,
>> Alexander
>>
>>
>
Received on Sat Sep 25 2010 - 18:55:11 CEST

This archive was generated by hypermail 2.2.0 : Mon Sep 27 2010 - 23:50:01 CEST