Segfault with rWishart in R 2.15.1

I just upgraded to R 2.15.1, and I am getting a segmentation fault when using the rWishart function (from the stats package) to sample moderately-size matrices.

Here is the output when I run R within gdb.  720 appears to be the dimensionality cut-off.  Anything smaller works fine.  Anything larger crashes.

x=rWishart(1,700,diag(700)) // this works
x=rWishart(1,720,diag(720))

Error: C stack usage is too close to the limit

x=rWishart(1,722,diag(722))

I just upgraded to R 2.15.1, and I am getting a segmentation fault when using the rWishart function (from the stats package) to sample moderately-size matrices.

Here is the output when I run R within gdb.  720 appears to be the dimensionality cut-off.  Anything smaller works fine.  Anything larger crashes.

x=rWishart(1,700,diag(700)) // this works
x=rWishart(1,720,diag(720))

Error: C stack usage is too close to the limit

x=rWishart(1,722,diag(722))

Program received signal EXC_BAD_ACCESS, Could not access memory.
Reason: KERN_PROTECTION_FAILURE at address: 0x00007fff5f3fff88
0x00007fff88b9c68b in __vfprintf ()
(gdb) bt
#0  0x00007fff88b9c68b in __vfprintf ()
#1  0x00007fff88b9c18e in vfprintf_l ()
#2  0x00007fff88ba52d0 in printf ()
#3  0x0000000100198619 in Rstd_WriteConsoleEx (
     buf=<value temporarily unavailable, due to optimizations>,
     len=<value temporarily unavailable, due to optimizations>,
     otype=<value temporarily unavailable, due to optimizations>)
     at ../../../../R-2.15.1/src/unix/sys-std.c:963
#4  0x00000001001497b7 in REvprintf (format=0x100253820 "%s",
     arg=0x7fff5f4029f0) at ../../../../R-2.15.1/src/main/printutils.c:813
#5  0x0000000100149a02 in REprintf (
     format=<value temporarily unavailable, due to optimizations>)
     at ../../../../R-2.15.1/src/main/printutils.c:698
#6  0x0000000100099e45 in verrorcall_dflt (call=0x10080db78,
     format=0x10025b9d8 "C stack usage is too close to the limit",
     ap=0x7fff5f408cc0) at ../../../../R-2.15.1/src/main/errors.c:656
#7  0x000000010009a86e in Rf_errorcall (call=0x10080db78,
     format=0x10025b9d8 "C stack usage is too close to the limit")
     at ../../../../R-2.15.1/src/main/errors.c:698
#8  0x000000010009c95c in R_CheckStack ()
     at ../../../../R-2.15.1/src/main/errors.c:93
#9  0x0000000102704e80 in R_rWishart ()
#10 0x000000010007af0c in do_dotcall (call=0x102f63710,
---Type <return> to continue, or q <return> to quit---
     op=<value temporarily unavailable, due to optimizations>,
     args=<value temporarily unavailable, due to optimizations>,
     env=<value temporarily unavailable, due to optimizations>)
     at ../../../../R-2.15.1/src/main/dotcode.c:543
#11 0x00000001000a128f in bcEval (
     body=<value temporarily unavailable, due to optimizations>,
     rho=0x103df1140,
     useCache=<value temporarily unavailable, due to optimizations>)
     at ../../../../R-2.15.1/src/main/eval.c:4442
#12 0x00000001000aa1f5 in Rf_eval (e=0x102f63748, rho=0x103df1140)
     at ../../../../R-2.15.1/src/main/eval.c:397
#13 0x00000001000af5d1 in Rf_applyClosure (call=0x103df0b58, op=0x102f62470,
     arglist=0x103df0df8, rho=0x10085b6a8, suppliedenv=0x10085b6e0)
     at ../../../../R-2.15.1/src/main/eval.c:859
#14 0x00000001000aa402 in Rf_eval (e=0x103df0b58, rho=0x10085b6a8)
     at ../../../../R-2.15.1/src/main/eval.c:510
#15 0x00000001000ad705 in do_set (call=0x103df0c38, op=0x100832030,
     args=0x103df0c00, rho=0x10085b6a8)
     at ../../../../R-2.15.1/src/main/eval.c:1715
#16 0x00000001000aa4ec in Rf_eval (e=0x103df0c38, rho=0x10085b6a8)
     at ../../../../R-2.15.1/src/main/eval.c:466
#17 0x00000001000e3822 in Rf_ReplIteration (rho=0x10085b6a8, savestack=0,
     browselevel=<value temporarily unavailable, due to optimizations>,
---Type <return> to continue, or q <return> to quit---
     state=0x7fff5fbfe7d0) at ../../../../R-2.15.1/src/main/main.c:256
#18 0x00000001000e3b11 in R_ReplConsole (rho=0x10085b6a8, savestack=0,
     browselevel=0) at ../../../../R-2.15.1/src/main/main.c:305
#19 0x00000001000e4020 in run_Rmainloop ()
     at ../../../../R-2.15.1/src/main/main.c:987
#20 0x0000000100000eeb in main ()

Here is my sessionInfo() output.  I am running R on a brand new (as in, yesterday) Macbook Pro.  This is the CRAN binary version.  I did not link to an optimized BLAS on this computer (yet).  However, I did try this on my office computer (Mac Pro) that I compiled from source, linking to Intel MKL, and that also crashes around k=722.

sessionInfo()

R version 2.15.1 (2012-06-22)
Platform: x86_64-apple-darwin9.8.0/x86_64 (64-bit)

locale:
[1] en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base

Anyway, since R is segfaulting on a brand new computer, I thought you would like to know.

Michael=

On Jul 4, 2012, at 03:12 , Michael Braun wrote:

I just upgraded to R 2.15.1, and I am getting a segmentation fault
when using the rWishart function (from the stats package) to sample
moderately-size matrices.

Here is the output when I run R within gdb.  720 appears to be the
dimensionality cut-off.  Anything smaller works fine.  Anything
larger crashes.

x=rWishart(1,700,diag(700)) // this works
x=rWishart(1,720,diag(720))

Error: C stack usage is too close to the limit

x=rWishart(1,722,diag(722))

That'll be due to allocating memory off the C stack with alloca().
Please file a bug report on this (nothing to do with Mac, much less
with the age of your computer).

In a tight spot, you can run R from the terminal after raising the C
stack limit with, say, "ulimit -s 32768", but of course the trouble
returns at twice the matrix size.

Notice that there are really two issues here. One is that the
R_CheckStack() safeguards are failing. The other is why we're using
stack allocation in the first place.