debugging R-devel on WinNT (was RE: [Rd] zlib location)

Liaw, Andy andy_liaw@merck.com
Mon, 26 Nov 2001 16:36:55 -0500


From: Prof Brian Ripley [mailto:ripley@stats.ox.ac.uk]

[...]

> Also, AFAIK no one is working on the Windows port at present, so the build
> of R-devel is only tested irregularly.  I expect some work will be done on
> it once 1.4.0 goes into feature freeze, but can't guarantee even that.

I had been trying to find out why La.svd on moderately large matrices (say
500 x 100) crash R-devel linked against ATLAS on WinNT.  Could some one
provide some hint on how to go about doing this?  I tried the instruction in
rw-faq, but keep getting stuck as follows:

(gdb) break WinMain
Breakpoint 1 at 0x4013c6: file ../graphapp/graphappmain.c, line 94.

(gdb) run
"/cygdrive/c/WINNT/system32/KERNEL32.dll": error reading line numbers

"/cygdrive/c/WINNT/system32/ADVAPI32.dll": error reading line numbers

"/cygdrive/c/WINNT/system32/VERSION.dll": error reading line numbers

"/cygdrive/c/WINNT/system32/LZ32.dll": error reading line numbers


Breakpoint 1, WinMain (Instance=0x400000, PrevInstance=0x0, CmdLine=0x232b0b
"", CmdShow=10) at ../graphapp/graphappmain.c:94

Starting program:
/cygdrive/c/Rtools/R-devel1105/src/gnuwin32/../../bin/Rgui.exe 

(gdb) break R_ReadConsole
Error: Cannot access memory at address 0x1f9a4

I tried adding the directories where several source files are (e.g.,
src/main, src/gnuwin32, src/gnuwin32/graphapp, and src/gnuwin32/front-ends),
but still get the same thing.

Also, as R starts, I get a bunch of dialog boxes that complained about DLL
relocated.  Is this normal?

If I just type "c" and run Rgui, and issue the command
"La.svd(matrix(runif(5e4), 5e2, 1e2))", I get a few more dialog boxes, one
saying "Invalid address specified to RtlFreeHeap(9f0000, 1efde80)", another
saying "HEAP(Rgui.exe)", and a SIGTRAP is generated.

I did one more (crazy? stupid?) thing:  I replaced the lapack.c in R-devel
(dated Nov. 14, 2001) with the one in R-1.3.1 *as is*.  It actually
compiled!  However, the result is the same.

Any help on how to get gdb to work is greatly appreciated!!

Regards,
Andy

-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-devel mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To: r-devel-request@stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._