[BioC] Rsubread crashes in 32bit linux

[Wei Shi]

Dear Robert,

That's great. Thanks for letting us know.

Cheers,
Wei

> Dear Wei,
>
> it also works on my side, the 32 bit machines do not crash anymore.
>
> thanks for solving this so quickly!
> robert.
>
> On 06/08/2012 04:10 AM, Wei Shi wrote:
>> Dear Robert, Dan and Peter,
>>
>> We have made changes to a number of functions in the package to reduce
>> the memory allocated to Rsubread by the operating system when it was
>> loaded. The new version has been committed to both bioc release
>> (Rsubread 1.6.4) and bioc devel (Rsubread 1.7.4). They should be
>> available to you in a day or two.
>>
>> Also, the buildindex() function no longer needs the allocation of 1GB
>> continuous memory region. But it will still consume at least 1GB of
>> memory when it is running, no matter what the given value of the
>> 'memory' parameter is.
>>
>> We have tested the new version on our 32-bit VM machine (it has 3GB of
>> memory and the value of 'memory' parameter used by buildindex was 2500)
>> and it solves all the reported problems, so we are pretty happy with it.
>> I hope the new version works in your computers/laptops, but please do
>> let us know if it doesn't.
>>
>> Sorry about the problems you have encountered. It's always a challenge
>> to develop a R package with so much C code in it!
>>
>> Cheers,
>> Wei
>>
>>
>>
>> On Jun 6, 2012, at 9:08 PM, Dan Du wrote:
>>
>>> Dear Wei,
>>>
>>> Here is a standard bioclite update, I think it is at the last step when
>>> compiling Rsubread.so, the memory usage exceeds 5.5g, then system
>>> freeze
>>> and I have to call it off. Same result when runing 'R CMD INSTALL
>>> Rsubread_1.6.3.tar.gz' from shell, or manually compile all .c file and
>>> run the last gcc statement. I guess there might just be a minimum ram
>>> requirement somewhere higher than 6g... I will do some more poking when
>>> I have time.
>>>
>>> 'gcc -std=gnu99 -shared -o Rsubread.so R_wrapper.o SNP_calling.o
>>> aligner.o atgcContent.o detectionCall.o detectionCallAnnotation.o
>>> exon-algorithms.o exon-align.o fullscan.o gene-algorithms.o
>>> gene-value-index.o hashtable.o index-builder.o input-files.o
>>> processExons.o propmapped.o qualityScores.o readSummary.o
>>> removeDuplicatedReads.o sam2bed.o sorted-hashtable.o -lpthread
>>> -DMAKE_FOR_EXON -L/usr/lib64/R/lib -lR'
>>>
>>> Also down there are the sessionInfo and full gcc version, please let me
>>> know if you need more information.
>>>
>>> Regards,
>>> Dan
>>> --------------------------------------------------------------------
>>>> source('http://www.bioconductor.org/biocLite.R')
>>>> biocLite('')
>>> BioC_mirror: http://bioconductor.org
>>> Using R version 2.15, BiocInstaller version 1.4.6.
>>> Installing package(s) ''
>>> Old packages: 'Rsubread'
>>> Update all/some/none? [a/s/n]: a
>>> trying URL
>>> 'http://www.bioconductor.org/packages/2.10/bioc/src/contrib/Rsubread_1.6.3.tar.gz'
>>> Content type 'application/x-gzip' length 21891723 bytes (20.9 Mb)
>>> opened URL
>>> ==================================================
>>> downloaded 20.9 Mb
>>>
>>> WARNING: ignoring environment value of R_HOME
>>> * installing *source* package ‘Rsubread’ ...
>>> ** libs
>>> gcc -std=gnu99 -I/usr/share/R/include -DNDEBUG     -DMAKE_FOR_EXON
>>> -fpic
>>> -O3 -pipe  -g  -c R_wrapper.c -o R_wrapper.o
>>> gcc -std=gnu99 -I/usr/share/R/include -DNDEBUG     -DMAKE_FOR_EXON
>>> -fpic
>>> -O3 -pipe  -g  -c SNP_calling.c -o SNP_calling.o
>>> gcc -std=gnu99 -I/usr/share/R/include -DNDEBUG     -DMAKE_FOR_EXON
>>


______________________________________________________________________
The information in this email is confidential and intend...{{dropped:4}}