Re: [Bioc-devel] caught segfault when compiling ensembldb

2019-10-17 Thread Martin Morgan
It's hard to tell from the stack trace, but is this reproducible in a new R (devel) session with loadNamespace("ensembldb") ? And did you actually install new packages in R-devel, or only re-use packages installed in an earlier version of R? If the latter, then definitely you'll want to re-i

Re: [Bioc-devel] caught segfault when compiling ensembldb

2019-10-17 Thread Martin Morgan
A preliminary insight is that it looks like you're using R-devel (aka R-4.0), whereas at the moment (until the next Bioconductor release in a couple of weeks...) the 'supported' devel version of Bioconductor uses R-3.6.*. Martin On 10/17/19, 10:10 AM, "Bioc-devel on behalf of Zhu, Jack (NIH/NC

[Bioc-devel] caught segfault when compiling ensembldb

2019-10-17 Thread Zhu, Jack (NIH/NCI) [E] via Bioc-devel
Dear all, I have encountered a very frustrating 'memory not mapped' problem when compiling ensembldb package: >BiocManager::install("ensembldb", ask = FALSE) Bioconductor version 3.10 (BiocManager 1.30.8), R Under development (unstable) (2019-10-15 r77295) Installing package(s) 'ensembldb' War