May 2007 - Darwin-dev - lists.apple.com

Re: [darwin-dev] Memory Allocation in Interrupt Handlers?
by Shawn Erickson 09 May '07

09 May '07

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com On May 8, 2007, at 7:39 AM, Alexander von Below wrote: Hello List, You should make all attempts to allocate memory before hand. -Shawn _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… the answer to this is probably easy to find in the right place, but is there a memory allocation routine that is interrupt safe? From the documentation for IOMalloc it is clear that it is not, so is there a function that is? Or would I have to allocate all my memory prior to going into interrupt context? This email sent to site_archiver(a)lists.apple.com

1 0

Re: catching floating point exception on OS X Intel
by Jonas Maebe 09 May '07

09 May '07

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com On 9 mei 2007, at 13:51, Jonas Maebe wrote: int setfpumask(int fpmask) { int oldmask; asm volatile ("fstcw _fpucw"); oldmask = fpucw & 0xffc0; fpucw = fpucw | (fpmask & FPU_ExceptionMask); Small error, that should read: int setfpumask(int fpmask) { int oldmask; asm volatile ("fstcw _fpucw"); oldmask = fpucw & ~0xffc0; fpucw = (fpucw & 0xffc0) | (fpmask & FPU_ExceptionMask); Jonas _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… This email sent to site_archiver(a)lists.apple.com

1 0

Re: catching floating point exception on OS X Intel
by Jonas Maebe 09 May '07

09 May '07

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com On 9 mei 2007, at 01:29, Kristoffer Eriksen wrote: Jonas #include <stdio.h> #include <stdlib.h> #include <math.h> #include <signal.h> #include <sys/types.h> #include <unistd.h> #define FPU_NoMask 0x00 #define FPU_Invalid 0x01 #define FPU_Denormal 0x02 #define FPU_DivisionByZero 0x04 #define FPU_Overflow 0x08 #define FPU_Underflow 0x10 #define FPU_Inexact 0x20 #define FPU_ExceptionMask 0xff unsigned short fpucw; unsigned int mxcsr; int setfpumask(int fpmask) { int oldmask; asm volatile ("fstcw _fpucw"); oldmask = fpucw & 0xffc0; fpucw = fpucw | (fpmask & FPU_ExceptionMask); asm volatile ("fldcw _fpucw"); asm volatile("fwait"); // the sse masking flags are the same as the 80x87 flags, just // shifted left 7 asm volatile("stmxcsr _mxcsr"); mxcsr = (mxcsr & 0xffffe07f) | ((fpmask & FPU_ExceptionMask) << 7); asm volatile("ldmxcsr _mxcsr"); return oldmask; } void dosqrtsse() { double x; pid_t pid; pid = getpid(); x = sqrt(-1.0/(double)pid); printf("sse: %d %f\n",pid,x); } void dosqrtx87() { long double x; pid_t pid; pid = getpid(); x = sqrt(-1.0/(long double)pid); printf("x87: %d %f\n",pid,x); } int main(int argc, char *argv[]){ printf("Default mask\n"); dosqrtsse(); dosqrtx87(); // both inexact and invalid need to be masked in this case printf("Only inexact and invalid exceptions masked\n"); setfpumask(FPU_Inexact|FPU_Invalid); dosqrtsse(); dosqrtx87(); printf("No exceptions masked\n"); setfpumask(FPU_NoMask); dosqrtsse(); // not reached dosqrtx87(); // not reached exit(0); } _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… To summarize, I'm using the feraiseexcept() function from fenv.h, which I had hoped would enable floating point exceptions for the requested flags. I figured you'd have to do something like this, but it doesn't work either: retval = fegetexceptflag(&flags, FE_INVALID | FE_DIVBYZERO | FE_OVERFLOW ); flags |= (FE_INVALID | FE_DIVBYZERO | FE_OVERFLOW); retval = fesetexceptflag(&flags, FE_INVALID | FE_DIVBYZERO | FE_OVERFLOW ); It's quite annoying that none of the man pages for those functions available on the Internet contain any example code (and that there aren't even any man pages for them in Mac OS X) Anyway, the following does work (based on code from the Free Pascal runtime library), but with a some caveats: a) I don't know how to write PIC inline assembler code in gcc (nor how to address local variables in inline gcc assembler, nor how to force local variables to be in memory), so this code will not work inside a dynamic library (because it directly accesses a global variable from assembler code) b) do not write GUI programs with fpu exceptions turned on, because there are parts of CoreGraphics which trigger such exceptions (at least on ppc) Float and double are handled using sse by default on Mac OS X, while long double is handled using the 80x87. This means you need to set both the sse and 80x87 fpu masks. You can force all code to use the 80x87 by compiling with -mfpmath=387 (not recommended, since slower). This email sent to site_archiver(a)lists.apple.com

1 0

Re: catching floating point exception on OS X Intel
by Alexey V. Medvedev 09 May '07

09 May '07

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Organization: GDT Software Group Hi, Kristoffer Eriksen wrote: > It correctly generates the nan, but doesn't crash. For kicks, if I > raise a SIGFPE manually, it crashes like it's supposed to, meaning > that the default behavior for a SIGFPE is correct. AFAIR, matherr is responsible for SIGFPE generation in sqrt() and other math library functions, try to dig for it. Try to replace sqrt() with simple division-by-zero instruction -- it'll be easier to make it work I think. Hope it helps -- Regargds, Alexey _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… This email sent to site_archiver(a)lists.apple.com

1 0

catching floating point exception on OS X Intel
by Kristoffer Eriksen 08 May '07

08 May '07

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Hi- #include <stdio.h> #include <stdlib.h> #include <math.h> #include <fenv.h> #include <signal.h> #include <sys/types.h> #include <unistd.h> int main(int argc, char *argv[]){ double x; int retval; pid_t pid; retval = -1; retval = feraiseexcept( FE_INVALID | FE_DIVBYZERO | FE_OVERFLOW ); pid = getpid(); x = sqrt(-1.0/(double)pid); printf("%d %d %f\n",retval,pid,x); //raise(SIGFPE); exit(0); } If I compile with: gcc feexcept.c -lm -Wall --std=c99 damocles:~/Research/feexcept[134] a.out 0 7065 nan Thanks, Kris Eriksen _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… I tried this question on darwin-x86 to no avail, so I thought I would try here. I am writing a large numerical code as part of my PhD dissertation. After coding for Linux for quite a while, I recently bought a quad- core 3.0 GHz Xeon MacPro, and am having a horrible time getting my code to trap floating point exceptions. I'm compiling with a standard Unix Makefile rather than Xcode, and an experimental GCC 4.3 release, which I need for its OpenMP support, though I get the same behavior with the standard /usr/bin/gcc (4.0.1). I am using OS X 10.4.8. For example, I would like this code to crash on an invalid floating point operation: To summarize, I'm using the feraiseexcept() function from fenv.h, which I had hoped would enable floating point exceptions for the requested flags. I get no compile errors or warnings. Unfortunately, the code runs to the end: It correctly generates the nan, but doesn't crash. For kicks, if I raise a SIGFPE manually, it crashes like it's supposed to, meaning that the default behavior for a SIGFPE is correct. I've found some examples about getting this to work on older Macs, but they all involve PPC assembly, which isn't an option here. I'm a scientist, not a systems programmer, so I may just be doing something stupid here. Any ideas? This email sent to site_archiver(a)lists.apple.com

1 0

[darwin-dev] Memory Allocation in Interrupt Handlers?
by Alexander von Below 08 May '07

08 May '07

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Hello List, Thanks a lot Alex _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… the answer to this is probably easy to find in the right place, but is there a memory allocation routine that is interrupt safe? From the documentation for IOMalloc it is clear that it is not, so is there a function that is? Or would I have to allocate all my memory prior to going into interrupt context? This email sent to site_archiver(a)lists.apple.com

1 0

Re: IOPlatformDevice vs. IODTPlatformDevice
by apic | puredarwin.org 08 May '07

08 May '07

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… This email sent to site_archiver(a)lists.apple.com

1 0

Re: IOPlatformDevice vs. IODTPlatformDevice
by Alexander von Below 08 May '07

08 May '07

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Hello apic, Alex Am 08.05.2007 um 14:31 schrieb apic | puredarwin.org: Hello everybody, thanks in advance and kind regards, apic puredarwin.org _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… you are asking this question a second time within a little over twelve hours. I will give this the benefit of the doubt and attribute it to a technical error. I can possibly see your frustration, because a search of "IODTPlatformDevice" turns up a gigantic count of zero results. Therefore I must say, I have no idea as to even what you are talking about. I hoped to find someone here who could answer me a question regarding IOKit, specificly IOPlatformDevice and IODTPlatformDevice. I would be happy if someone could explain me the diffrence and what changed during the time to it. I was specificly looking at the IOACPIFamily.kext and saw one fo the main changes from 3 to ver. 6 was that this superclass was changed within it. Is it possible to explain me why? This email sent to site_archiver(a)lists.apple.com

1 0

(no subject)
by apic | puredarwin.org 08 May '07

08 May '07

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hello everybody, thanks in advance and kind regards, apic puredarwin.org -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.7 (Darwin) iD8DBQFGQG23BDxoGUAtYxoRAlHsAJ9WWy/r0ydjGCoEaZRNTX8/hPc4GACeNFah Q5qZDlncI5rj2R4W3lRyF6E= =e7ni -----END PGP SIGNATURE----- _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… I hoped to find someone here who could answer me a question regarding IOKit, specificly IOPlatformDevice and IODTPlatformDevice. I would be happy if someone could explain me the diffrence and what changed during the time to it. I was specificly looking at the IOACPIFamily.kext and saw one fo the main changes from 3 to ver. 6 was that this superclass was changed within it. Is it possible to explain me why? This email sent to site_archiver(a)lists.apple.com

1 0

Detecting gdb.
by Steve Checkoway 08 May '07

08 May '07

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com -- Steve Checkoway _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… This email sent to site_archiver(a)lists.apple.com Currently, I'm using <http://developer.apple.com/qa/qa2004/ qa1361.html> to detect if I'm running under gdb conditionalized on debug builds as per the comment. I only use it to detect if gdb is running in which case I call DebugStr() instead of performing the usual crash handler code for failed asserts. (We treat a failed assert as crash, call the crash handler and dump the stack and all relevant data.) My question is since I am building using the 10.2.8 sdk since I haven't had any reason to upgrade that, how will I know if this changes apart from things going horribly wrong in debug builds? Will the headers change in the sdk to reflect any changes? My guess is no since code compiled against that sdk is supposed to run on that OS. Or am I worrying (and not very hard at that) about something that isn't likely to change for years? smime.p7s

1 0