[Rd] Plot dies with memory not mapped (segfault) (PR#9785)
Full_Name: Clay B
Version: 2.5.0 (2007-04-23)
OS: Solaris Nevada Build 55b
Submission from: (NULL) (65.101.229.198)
I find that running this script causes R to segfault reliably. However, running
just for one system at a time (modifying the for loop updating iter to run just
for a system at a time works). The system is a Sun W2100z with 12 GB of ram, and
R segfaults using only around 360 MB of ram. I can include MySQL output fed to
the script on request, but it's usually for failure data around a max of 1,000
data points in the "failureQuery", and less than 30,000 for "usageQuery". I can
generate a core if desired too.
begin traceback
--
s> source("SysVUserTime.R")
Loading required package: DBI
*** caught segfault ***
address f9, cause 'memory not mapped'
Traceback:
1: axis(side = side, at = at, labels = labels, ...)
2: Axis.default(...)
3: Axis(...)
4: localAxis(x, side = 1, ...)
5: plot.default(g$fit, g$res, xlab = "Fitted", ylab = "Residuals")
6: plot(g$fit, g$res, xlab = "Fitted", ylab = "Residuals")
7: eval.with.vis(expr, envir, enclos)
8: eval.with.vis(ei, envir)
9: source("SysVUserTime.R")
Possible actions:
1: abort (with core dump, if enabled)
2: normal R exit
3: exit R without saving workspace
4: exit R saving workspace
Selection: 3
- end traceback
---
begin script
-
#Connects to database
library(RMySQL)
library(chron)
CLAY_CLASS<-TRUE
PIE_CHARTS<-FALSE
BOXPLOTS<-TRUE
PLOTTING<-TRUE
OUTPUT_TEXT<-FALSE
ACF_PLOT<-TRUE
query_dates<-c("Year","Quarter","Month","Week")
oldpar <- par(no.readonly = TRUE)
drv <- dbDriver("MySQL")
con <- dbConnect(drv, user="macs370",dbname="lanl",password="macs370")
systems<-c("s8_usage_clean","s15_usage_clean","s16_usage_clean","s23_usage_clean","s20_usage_node_num_cpu")
systemNum<-c("8","15","16","23","20")
# Query for each system
for (iter in 1:length(systems)) {
#for (iter in 5:5) {
rm(dat,res,IQR_Stats,userStatTimes,cors,data)
if(OUTPUT_TEXT!=TRUE)
{
sink(paste("/export/MACS370_project/Image/SysVUserTime/CorrelationsSys",systemNum[iter],".TXT",sep=""))
}
if(PIE_CHARTS==TRUE) {
for (iter in 1:length(systems)) {
X11()
res<- dbSendQuery(con, paste("SELECT
SUM(cpu_seconds_user)/SUM(total_time), SUM(cpu_seconds_system)/SUM(total_time)
FROM ",systems[iter],";",sep=""))
dat<- fetch(res,n=-1)
data<-rbind(dat[,1],dat[,2])
pie(data,labels=c("User","System"),main=paste("System
vs. User Time for System ",systemNum[iter],"by",QUERY_DATE,sep=""))
}
}
# Query for each date segmentation
for(k in 1:length(query_dates)) {
#for(k in 4:4) {
#for(k in 2:2) {
QUERY_DATE<-query_dates[k]
userTime<-NULL
failures<-NULL
# Date Clamp Limits Us To The Middle 80% Of All Data
usageDateClamp<-paste(" WHERE FROM_UNIXTIME(submit_time) >= (SELECT ",
"FROM_UNIXTIME(MIN(submit_time)) + INTERVAL 0.10*",
"DATEDIFF(MAX(FROM_UNIXTIME(submit_time)),MIN(FROM_UNIXTIME(",
"submit_time))) DAY FROM ",systems[iter],") AND
FROM_UNIXTIME(",
"submit_time) <= (SELECT FROM_UNIXTIME(MAX(submit_time)) - ",
"INTERVAL 0.10*DATEDIFF(MAX(FROM_UNIXTIME(submit_time)),MIN(",
"FROM_UNIXTIME(submit_time))) DAY FROM ",systems[iter],") ",
sep="")
failureDateClamp<-paste(" AND prob_start_date >= (SELECT ",
"FROM_UNIXTIME(MIN(submit_time)) + INTERVAL 0.10*",
"DATEDIFF(MAX(FROM_UNIXTIME(submit_time)),MIN(FROM_UNIXTIME(",
"submit_time))) DAY FROM ",systems[iter],") ",
"AND prob_start_date <= (SELECT ",
"FROM_UNIXTIME(MAX(submit_time)) - INTERVAL 0.10*",
"DATEDIFF(MAX(FROM_UNIXTIME(submit_time)),MIN(FROM_UNIXTIME(",
"submit_time))) DAY FROM ",systems[iter],") ",sep="")
if(QUERY_DATE=="Year") {
usageAVGQuery<-paste("SELECT
STR_TO_DATE(CONCAT(YEAR(FROM_UNIXTIME(dispatch_time)),'-01-01'),'%Y-%m-%d') AS
date, AVG(cpu_seconds_user/total_time) AS time, AVG(num_cpus) AS AVG_CPUs,
AVG(dispatch_time-end_time) AS AVG_job_length FROM
",systems[iter],usageDateClamp,"GROUP BY date;",sep="")
usageQuery<-paste("SELECT
STR_TO_DATE(CONCAT(YEAR(FROM_UNIXTIME(dispatch_time)),'-01-01'),'%Y-%m-%d') AS
date, cpu_seconds_user/total_time AS time FROM ",systems[iter],"
",usageDateClamp,";",sep="")
if(CLAY_CLASS=="TRUE") {
failureQuery<-paste("SELECT
DISTINCT(STR_TO_DATE(CONCAT(YEAR(FROM_UNIXTIME(dispatch_time)),'-01-01'),'%Y-%m-%d'))
AS date, if(tbl.probs is NULL,0,tbl.probs) FROM ",systems[iter]," LEFT JOIN
(SELECT SUM(if(clay_class
[Rd] How to preserve data across function calls in a library package
Hi,
I am writing an R package with two functions in C++. So far
everything works.
Now, i would like to write a third function which would use a pointer
(it is a pointer to a class object) created by first function.
I tried placing this pointer outside of the function definitions
(i.e to make it global) but when called in the 3rd function i get
>*** caught bus error ***
address 0x0, cause 'invalid alignment'"
I tried Callocing it in the 1st function but to no avail. Here is a
quick summary. When foo is called (through do_foo, **after** having
called do_kNN_e) i get the aforementioned error.
Can anyone provide some pointers (no pun intended) on this?
Thanks
Saptarshi
ANN* book;
int* foot;
void foo(void){
Rprintf("many times\n");
Rprintf("%p\n",book);
Rprintf("%p\n",foot);
}
SEXP
kNN_e(SEXP data, SEXP Nrow, SEXP Ncol,SEXP K,SEXP Eps)
{
int nrow=asInteger(Nrow);
int ncol=asInteger(Ncol);
int k=asInteger(K);
double eps=asReal(Eps);
SEXP ans,distance;
SEXP retlist;
PROTECT(ans=allocMatrix(INTSXP,nrow,k)); //The 2nd argument gives
the number of rows, and the last the number of cols see http://cran.r-
project.org/doc/manuals/R-exts.html
PROTECT(distance=allocMatrix(REALSXP,nrow,k));
ANNpointArray datapoints;
ANNpoint qpoint;
ANNkd_tree* kdTree;
book=Calloc(1,ANN*);
foot=Calloc(1,int);
book=kdTree;
*foot=10;
...
}
extern "C" {
void do_foo(void){
foo();
}
SEXP
do_kNN_e(SEXP data, SEXP Nrow, SEXP Ncol,SEXP k,SEXP eps)
{
return kNN_e(data,Nrow, Ncol,
k,eps);
}
[[alternative HTML version deleted]]
__
R-devel@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel
Re: [Rd] [R] How to preserve data across function calls in a library package
Hi,
Some progress: I am using
SEXP retty;
book=Calloc(1,int);
*book=10;
PROTECT(retty=R_MakeExternalPtr(book,R_NilValue,R_NilValue));
then UNPROTECTING and returning retty.
In a another function,
foo(SEXP s){
int* f=(int *)R_ExternalPtrAddr(p);
Rprintf("many times %d\n",*f);
}
When called do_foo(p) where do_foo calls foo and p is the pointer
returned by the former code snippet, the Rprintf successfully prints
the correct value but subsequently crashes
*** caught bus error ***
address 0x0, cause 'invalid alignment'.
I can't figure out why... I would appreciate any advice provided.
Rgds
Saptarshi
On Jul 10, 2007, at 11:41 AM, Saptarshi Guha wrote:
> Hi,
> I am writing an R package with two functions in C++. So far
> everything works.
> Now, i would like to write a third function which would use a pointer
> (it is a pointer to a class object) created by first function.
> I tried placing this pointer outside of the function definitions
> (i.e to make it global) but when called in the 3rd function i get
>> *** caught bus error ***
> address 0x0, cause 'invalid alignment'"
>
> I tried Callocing it in the 1st function but to no avail. Here is a
> quick summary. When foo is called (through do_foo, **after** having
> called do_kNN_e) i get the aforementioned error.
> Can anyone provide some pointers (no pun intended) on this?
>
> Thanks
> Saptarshi
>
> ANN* book;
> int* foot;
>
> void foo(void){
>Rprintf("many times\n");
>Rprintf("%p\n",book);
>Rprintf("%p\n",foot);
> }
>
> SEXP
> kNN_e(SEXP data, SEXP Nrow, SEXP Ncol,SEXP K,SEXP Eps)
> {
>int nrow=asInteger(Nrow);
>int ncol=asInteger(Ncol);
>int k=asInteger(K);
>double eps=asReal(Eps);
>
>SEXP ans,distance;
>SEXP retlist;
>PROTECT(ans=allocMatrix(INTSXP,nrow,k)); //The 2nd argument gives
> the number of rows, and the last the number of cols see http://cran.r-
> project.org/doc/manuals/R-exts.html
>PROTECT(distance=allocMatrix(REALSXP,nrow,k));
>ANNpointArray datapoints;
>ANNpoint qpoint;
>ANNkd_tree* kdTree;
>book=Calloc(1,ANN*);
>foot=Calloc(1,int);
>book=kdTree;
> *foot=10;
>
> ...
> }
>
> extern "C" {
>void do_foo(void){
> foo();
>}
>
> SEXP
> do_kNN_e(SEXP data, SEXP Nrow, SEXP Ncol,SEXP k,SEXP eps)
> {
>return kNN_e(data,Nrow, Ncol,
>k,eps);
>
> }
>
>
>
> [[alternative HTML version deleted]]
>
> __
> [EMAIL PROTECTED] mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide http://www.R-project.org/posting-
> guide.html
> and provide commented, minimal, self-contained, reproducible code.
Saptarshi Guha | [EMAIL PROTECTED] | http://www.stat.purdue.edu/~sguha
Would you people stop playing these stupid games?!?!?
__
R-devel@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel
Re: [Rd] BLAS / LAPACK version information from within R-session?
And Brian Ripley writes: > I know of no way to ask a BLAS dynamic library what version it > is. Tinkering with the BLAS interface hasn't fared so well, but LAPACK 3.1 added an ILAVER subroutine that returns the major, minor, and patch level versions for LAPACK itself. >> An obvious application could be safety-checks for packages >> like Matrix and quantreg at load / attach - time. > > Why should that be needed? To warn of bugs in specific LAPACK versions, perhaps. I know I let a couple of embarassing typos merge into 3.1.0 with the eigenvalue drivers. The ones I know of so far only affect performance, but there may be another that leads to an unreported error. Looks like R makes DSYEVR available through eigen.R, so it's affected. Sorry. And I know prior LAPACK versions have some painful error cases. They seem to impact few people, so having the LAPACK version in related bug reports may help remove those causes. >> Perhaps this information may even be included into the output >> of R.Version() :-) ? > > I can think of more worthy candidates, e.g. the compilers used > to build R, and for Linux something more precise about the > distro used. Alas, yes, those could be very useful with the resurgence of funny arithmetics. Jason __ R-devel@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-devel
[Rd] Getting param names of primitives
Hi, In the latest R this does not work anymore args <- formals(log) because log is primitive. Is there any other way to get the argument list? I just need the name of the parameter to display some error message about it. I have looked at args() but I can't find anything to extract the argument names from it. Thanks, M. Manese __ R-devel@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-devel
Re: [Rd] Getting param names of primitives
> formals(args(log)) $x $base exp(1) gives what formals(log) used to. On Wed, 11 Jul 2007, miguel manese wrote: > Hi, > > In the latest R this does not work anymore > > args <- formals(log) > > because log is primitive. Is there any other way to get the argument > list? I just need the name of the parameter to display some error > message about it. I have looked at args() but I can't find anything to > extract the argument names from it. > > Thanks, > M. Manese -- Brian D. Ripley, [EMAIL PROTECTED] Professor of Applied Statistics, http://www.stats.ox.ac.uk/~ripley/ University of Oxford, Tel: +44 1865 272861 (self) 1 South Parks Road, +44 1865 272866 (PA) Oxford OX1 3TG, UKFax: +44 1865 272595 __ R-devel@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-devel
Re: [Rd] Getting param names of primitives
Thanks! Worked like a charm. Regards, M. Manese On 7/11/07, Prof Brian Ripley <[EMAIL PROTECTED]> wrote: > > formals(args(log)) > $x > > > $base > exp(1) > > gives what formals(log) used to. > > On Wed, 11 Jul 2007, miguel manese wrote: > > > Hi, > > > > In the latest R this does not work anymore > > > > args <- formals(log) > > > > because log is primitive. Is there any other way to get the argument > > list? I just need the name of the parameter to display some error > > message about it. I have looked at args() but I can't find anything to > > extract the argument names from it. > > > > Thanks, > > M. Manese > > -- > Brian D. Ripley, [EMAIL PROTECTED] > Professor of Applied Statistics, http://www.stats.ox.ac.uk/~ripley/ > University of Oxford, Tel: +44 1865 272861 (self) > 1 South Parks Road, +44 1865 272866 (PA) > Oxford OX1 3TG, UKFax: +44 1865 272595 > __ R-devel@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-devel
