Is there any way to implement power management functionality using QNX 4.25
(Proc32=4.24). As we require the following functionality:
Get current battery status (PC104 based system).
Complete power down of system, without user intervention (PC104 based
system).
Is this possible, I have tried int15 calls but the code just seems to
sigsegv when ever it is run. Any help on this will be gratefully received.
Len.
Len Meakin <len.meakin@cirrustesting.com> wrote:
Is there any way to implement power management functionality using QNX 4.25
(Proc32=4.24). As we require the following functionality:Get current battery status (PC104 based system).
Complete power down of system, without user intervention (PC104 based
system).Is this possible, I have tried int15 calls but the code just seems to
sigsegv when ever it is run. Any help on this will be gratefully received.
QNX does not provide any power management functionality for QNX 4.
I don’t know if the int15 calls are callable from 32-bit protected mode,
nor how you went about trying to call them.
At minimum, you will need to link your program with -T1 and run as root.
QNX Training Services
http://www.qnx.com/support/training/
Please followup in this newsgroup if you have further questions.
You could throw hardware at the problem 
A product available from Tri-M:
http://www.tri-m.com/products/engineering/hescser.html
Tthis will allow you to monitor the status of input power,
and battery status.
You can code any actions you want to take.
Len Meakin wrote:
Is there any way to implement power management functionality using QNX 4.25
(Proc32=4.24). As we require the following functionality:Get current battery status (PC104 based system).
Complete power down of system, without user intervention (PC104 based
system).Is this possible, I have tried int15 calls but the code just seems to
sigsegv when ever it is run. Any help on this will be gratefully received.Len.
Already linked with T1 and running as root, so I guess direct int15 calls
are out of the question. I have found some apm code (Apm-1.3.tar.gz) on
QUICS, but this seems to only support QNX 4.25 (Proc4.25), but we have to
run with Proc4.24 because of our int10 usage. I’ll be taking a closer look
at this code today to determine if it will provide any hints on how to
accomplish this.
Thank you,
Len.
“David Gibbs” <dagibbs@qnx.com> wrote in message
news:bhis3c$5uq$2@nntp.qnx.com…
Len Meakin <> len.meakin@cirrustesting.com> > wrote:
Is there any way to implement power management functionality using QNX
4.25
(Proc32=4.24). As we require the following functionality:Get current battery status (PC104 based system).
Complete power down of system, without user intervention (PC104 based
system).Is this possible, I have tried int15 calls but the code just seems to
sigsegv when ever it is run. Any help on this will be gratefully
received.QNX does not provide any power management functionality for QNX 4.
I don’t know if the int15 calls are callable from 32-bit protected mode,
nor how you went about trying to call them.At minimum, you will need to link your program with -T1 and run as root.
-David
QNX Training Services
http://www.qnx.com/support/training/
Please followup in this newsgroup if you have further questions.
Thanks for the link, but space is very limited so I am guessing that a
hardware solution is not a real option at this time.
Thank you,
Len.
<acellarius@yahoo.com> wrote in message news:3F3D2D3C.2AF800C5@yahoo.com…
You could throw hardware at the problem >
A product available from Tri-M:
http://www.tri-m.com/products/engineering/hescser.htmlTthis will allow you to monitor the status of input power,
and battery status.
You can code any actions you want to take.Len Meakin wrote:
Is there any way to implement power management functionality using QNX
4.25
(Proc32=4.24). As we require the following functionality:Get current battery status (PC104 based system).
Complete power down of system, without user intervention (PC104 based
system).Is this possible, I have tried int15 calls but the code just seems to
sigsegv when ever it is run. Any help on this will be gratefully
received.Len.
If this is QNX4 there was an int386() function that will help you make
this kinf of call, I think. Check the docs.
Len Meakin <len.meakin@cirrustesting.com_remove> wrote:
LM > Already linked with T1 and running as root, so I guess direct int15 calls
LM > are out of the question. I have found some apm code (Apm-1.3.tar.gz) on
LM > QUICS, but this seems to only support QNX 4.25 (Proc4.25), but we have to
LM > run with Proc4.24 because of our int10 usage. I’ll be taking a closer look
LM > at this code today to determine if it will provide any hints on how to
LM > accomplish this.
LM > Thank you,
LM > Len.
LM > “David Gibbs” <dagibbs@qnx.com> wrote in message
LM > news:bhis3c$5uq$2@nntp.qnx.com…
Len Meakin <> len.meakin@cirrustesting.com> > wrote:
Is there any way to implement power management functionality using QNX
LM > 4.25
(Proc32=4.24). As we require the following functionality:Get current battery status (PC104 based system).
Complete power down of system, without user intervention (PC104 based
system).Is this possible, I have tried int15 calls but the code just seems to
sigsegv when ever it is run. Any help on this will be gratefully
LM > received.QNX does not provide any power management functionality for QNX 4.
I don’t know if the int15 calls are callable from 32-bit protected mode,
nor how you went about trying to call them.At minimum, you will need to link your program with -T1 and run as root.
-David
QNX Training Services
http://www.qnx.com/support/training/
Please followup in this newsgroup if you have further questions.
–
Bill Caroselli – Q-TPS Consulting
1-(626) 824-7983
qtps@earthlink.net
Tried that too, that is where the code sigsegvs: modified from _intr_v86 to
int386:
Linked with -T1 and run as root, the very first int386 call seems to be
where the problem lies (int15 call).
<===========================================
/*A modified apm.c previously posted here, just to power-down
(not just “it is now safe to reboot your computer”)
APM-capable machines.
(Didn’t test on ACPI-only machines)
compile:
make apmoff; usemsg apmoff apmoff.c
Now, how should I fire this after “shutdown” “phshutdown” ?
(daemonize, trap SIGPWR, wait 9 secs then powerdown?)
/
/
#define APM_SYSTEM_BIOS 0x15
#define APM_DEV_APM_BIOS 0x0000
#define APM_INSTALLATION_CHECK 0x5300
#define APM_32BIT_CONNECT 0x5303
#define APM_DISCONNECT 0x5304
#define APM_SET_POWER_STATE 0x5307
#define APM_GET_POWER_STATUS 0x530a
#define APM_DRIVER_VERSION 0x530e
#define APM_GET_CAPABILITIES 0x5310
#define PSL_C 1
#define BH(reg) ((((reg).x.ebx)>>8)&0xff)
#define BL(reg) (((reg).x.ebx)&0xff)
#define CH(reg) ((((reg).x.ecx)>>8)&0xff)
#define CL(reg) (((reg).x.ecx)&0xff)
#define DPRINTF(x) if (option.verbose) { printf x; }
#define DUMPREGS(x)
struct {
u_short apm_flags;
u_short apm_version; /* BCD xx.xx /
u_short apm_driver_version; / BCD xx.xx */
} apm_info;
#ifdef __USAGE
%C - Shutdown the power on APM platform
%C [-v] [-S off|standby|suspend]
-v Verbose
-S Next state (default: off)
#endif
struct {
int verbose;
int dryrun;
enum {Standby=1, Suspend=2, Off=3} setstate;
} option = {0,0,Off};
void dumpregs(const union REGS *reg)
{
DPRINTF((“AX=%04X BX=%04X CX=%04X DX=%04X\n”,
(int)reg->x.eax, (int)reg->x.ebx, (int)reg->x.ecx,
(int)reg->x.edx));
DPRINTF((“FL=%04X ESI=%08lX\n”, (int)reg->w.cflag, reg->x.esi));
}
int main(int argc, char *argv[])
{ union REGS reg;
int s;
int c;
while ((c = getopt(argc, argv, “vS:n”)) != -1)
{ switch (c)
{ case ‘v’:
option.verbose++;
break;
case ‘S’:
if (!strcmpi(optarg, “shutdown”) || !strcmpi(optarg, “off”))
{ option.setstate = Off;
}else if (!strcmpi(optarg, “suspend”))
{ option.setstate = Suspend;
} else if (!strcmpi(optarg, “standby”))
{ option.setstate = Standby;
} else
{ fprintf(stderr, “Unknown option -%c %s\n”, c, optarg);
exit(1);
}
break;
case ‘n’:
option.dryrun++;
break;
case ‘?’:
fprintf(stderr, “Unknown option -%c\n”, c);
exit(1);
}
}
/– Installation check -------------------------------------------/
DPRINTF((“APM: Installation Check\n”));
memset(®, 0, sizeof(reg));
reg.x.eax = APM_INSTALLATION_CHECK;
reg.x.ebx = APM_DEV_APM_BIOS;
s = int386(APM_SYSTEM_BIOS, ®, ®);
DPRINTF((“APM: Never Get here as we SIGSEGV (int386)\n”));
if (s || reg.x.cflag&PSL_C) { fprintf(stderr,“error on int386\n”);
exit(s); }
DUMPREGS(®);/XXX/
/*
1 has32bit
2 Idle slows CPU
3 disable
4 disengage
*/
DPRINTF((" APM Supported: %s\n", (reg.x.cflag&PSL_C)?“no”:“yes”));
apm_info.apm_version = reg.x.eax;
apm_info.apm_flags = reg.x.ecx;
DPRINTF((" BX signature=<%c%c>%s\n",
BH(reg), BL(reg),
((reg.x.ebx&0xffff)==(‘P’<<8)+‘M’)?" (ok)":" (MISMATCH)"));
DPRINTF((" BIOS APM version %x.%x\n",
(apm_info.apm_version>>8)&255, apm_info.apm_version&255 ));
if(apm_info.apm_flags & 1) { DPRINTF((" 16bit mode interface\n")); }
if(apm_info.apm_flags & 2) { DPRINTF((" 32bit mode interface\n")); }
if(apm_info.apm_flags & 4) { DPRINTF((" Slow CPU on Idle\n")); }
DPRINTF((" BIOS PM %s\n", (apm_info.apm_flags&8)?“disabled”:“enabled”));
DPRINTF((" BIOS PM %s\n",
(apm_info.apm_flags&16)?“disengaged”:“engaged”));
if ((reg.x.ebx&0xffff) != (‘P’<<8)+‘M’) {
fprintf(stderr, “APM is not supported on this machine.\n”);
exit(1);
}
/* disconnect first, in case somebody was connected */
memset(®, 0, sizeof(reg));
reg.x.eax = APM_DISCONNECT;
reg.x.ebx = APM_DEV_APM_BIOS;
s = int386(APM_SYSTEM_BIOS, ®, ®);
if (s) { fprintf(stderr,“error on int386\n”); exit(s); }
/* connect 32 /
DPRINTF((“APM: Connect32\n”));
memset(®, 0, sizeof(reg));
reg.x.eax = APM_32BIT_CONNECT;
reg.x.ebx = APM_DEV_APM_BIOS;
s = int386(APM_SYSTEM_BIOS, ®, ®);
if (s) { fprintf(stderr,“error on int386\n”); exit(s); }
if (reg.x.cflag&PSL_C) {
fprintf(stderr, “APM: Connect32 failed (code 0x%02lX)\n”, reg.x.eax);
return 1;
}
DUMPREGS(®);
/
16 code16 length-1
/* set my version (needed to access 1.1- BIOS calls) (APM 1.1) /
apm_info.apm_driver_version = 0x0101; / set to APM 1.1 /
DPRINTF((“APM: set Driver Version to %x.%x\n”,
(apm_info.apm_driver_version>>8)&255,
apm_info.apm_driver_version&255
));
memset(®, 0, sizeof(reg));
reg.x.eax = APM_DRIVER_VERSION;
reg.x.ebx = APM_DEV_APM_BIOS;
reg.x.ecx = apm_info.apm_driver_version;
s = int386(APM_SYSTEM_BIOS, ®, ®);
/
DPRINTF((“APM: Disconnect\n”));
memset(®, 0, sizeof(reg));
reg.x.eax = APM_DISCONNECT;
reg.x.ebx = APM_DEV_APM_BIOS; /APM BIOS/
s = int386(APM_SYSTEM_BIOS, ®, ®);
if (s) { fprintf(stderr,“error on int386\n”); exit(s); }
return s;
}
===========================================>
Thanks,
Len.
“Bill Caroselli” <qtps@earthlink.net> wrote in message
news:bhqinf$m90$2@inn.qnx.com…
If this is QNX4 there was an int386() function that will help you make
this kinf of call, I think. Check the docs.
Len Meakin <> len.meakin@cirrustesting.com> _remove> wrote:
LM > Already linked with T1 and running as root, so I guess direct int15
calls
LM > are out of the question. I have found some apm code (Apm-1.3.tar.gz)
on
LM > QUICS, but this seems to only support QNX 4.25 (Proc4.25), but we
have to
LM > run with Proc4.24 because of our int10 usage. I’ll be taking a closer
look
LM > at this code today to determine if it will provide any hints on how
to
LM > accomplish this.LM > Thank you,
LM > Len.LM > “David Gibbs” <> dagibbs@qnx.com> > wrote in message
LM > news:bhis3c$5uq$> 2@nntp.qnx.com> …
Len Meakin <> len.meakin@cirrustesting.com> > wrote:
Is there any way to implement power management functionality using
QNX
LM > 4.25
(Proc32=4.24). As we require the following functionality:Get current battery status (PC104 based system).
Complete power down of system, without user intervention (PC104 based
system).Is this possible, I have tried int15 calls but the code just seems to
sigsegv when ever it is run. Any help on this will be gratefully
LM > received.QNX does not provide any power management functionality for QNX 4.
I don’t know if the int15 calls are callable from 32-bit protected
mode,
nor how you went about trying to call them.At minimum, you will need to link your program with -T1 and run as
root.-David
QNX Training Services
http://www.qnx.com/support/training/
Please followup in this newsgroup if you have further questions.
\Bill Caroselli – Q-TPS Consulting
1-(626) 824-7983
qtps@earthlink.net
I’m still really stuck with this, has anyone got any ideas on how I
could overcome this problem ? Any suggestions will be welcomed (apart
from using a hardware solution as this is not possible).
Thank you,
Len.
Len Meakin wrote:
Tried that too, that is where the code sigsegvs: modified from _intr_v86 to
int386:Linked with -T1 and run as root, the very first int386 call seems to be
where the problem lies (int15 call).===========================================
/*A modified apm.c previously posted here, just to power-down
(not just “it is now safe to reboot your computer”)
APM-capable machines.
(Didn’t test on ACPI-only machines)compile:
make apmoff; usemsg apmoff apmoff.cNow, how should I fire this after “shutdown” “phshutdown” ?
(daemonize, trap SIGPWR, wait 9 secs then powerdown?)–
kabe/
/
- apmoff.c
- APM power shutdown
- for QNX RTP(x86)
/
/--------------------------------------------------------------------------
-/
/– Modified for use for QNX 4.25 (well attempt to
modify) -----------------/
/--------------------------------------------------------------------------
-*/
#include <stdio.h
#include <stdlib.h
#include <unistd.h
#include <sys/types.h
#include <string.h
#include <i86.h#define APM_SYSTEM_BIOS 0x15
#define APM_DEV_APM_BIOS 0x0000
#define APM_INSTALLATION_CHECK 0x5300
#define APM_32BIT_CONNECT 0x5303
#define APM_DISCONNECT 0x5304
#define APM_SET_POWER_STATE 0x5307
#define APM_GET_POWER_STATUS 0x530a
#define APM_DRIVER_VERSION 0x530e
#define APM_GET_CAPABILITIES 0x5310
#define PSL_C 1#define BH(reg) ((((reg).x.ebx)>>8)&0xff)
#define BL(reg) (((reg).x.ebx)&0xff)
#define CH(reg) ((((reg).x.ecx)>>8)&0xff)
#define CL(reg) (((reg).x.ecx)&0xff)#define DPRINTF(x) if (option.verbose) { printf x; }
#define DUMPREGS(x)struct {
u_short apm_flags;
u_short apm_version; /* BCD xx.xx /
u_short apm_driver_version; / BCD xx.xx */
} apm_info;
#ifdef __USAGE
%C - Shutdown the power on APM platform%C [-v] [-S off|standby|suspend]
-v Verbose
-S Next state (default: off)
#endifstruct {
int verbose;
int dryrun;
enum {Standby=1, Suspend=2, Off=3} setstate;
} option = {0,0,Off};void dumpregs(const union REGS *reg)
{
DPRINTF((“AX=%04X BX=%04X CX=%04X DX=%04X\n”,
(int)reg->x.eax, (int)reg->x.ebx, (int)reg->x.ecx,
(int)reg->x.edx));
DPRINTF((“FL=%04X ESI=%08lX\n”, (int)reg->w.cflag, reg->x.esi));
}int main(int argc, char *argv[])
{ union REGS reg;
int s;
int c;while ((c = getopt(argc, argv, “vS:n”)) != -1)
{ switch (c)
{ case ‘v’:
option.verbose++;
break;
case ‘S’:
if (!strcmpi(optarg, “shutdown”) || !strcmpi(optarg, “off”))
{ option.setstate = Off;
}else if (!strcmpi(optarg, “suspend”))
{ option.setstate = Suspend;
} else if (!strcmpi(optarg, “standby”))
{ option.setstate = Standby;
} else
{ fprintf(stderr, “Unknown option -%c %s\n”, c, optarg);
exit(1);
}
break;
case ‘n’:
option.dryrun++;
break;
case ‘?’:
fprintf(stderr, “Unknown option -%c\n”, c);
exit(1);
}
}/– Installation check -------------------------------------------/
DPRINTF((“APM: Installation Check\n”));
memset(®, 0, sizeof(reg));
reg.x.eax = APM_INSTALLATION_CHECK;
reg.x.ebx = APM_DEV_APM_BIOS;
s = int386(APM_SYSTEM_BIOS, ®, ®);
DPRINTF((“APM: Never Get here as we SIGSEGV (int386)\n”));
if (s || reg.x.cflag&PSL_C) { fprintf(stderr,“error on int386\n”);
exit(s); }
DUMPREGS(®);/XXX/
/*
- AH.AL: APM version number (BCD)
- BH.BL: “PM”
- CX: 0 has16bit
1 has32bit2 Idle slows CPU3 disable4 disengage*/
DPRINTF((" APM Supported: %s\n", (reg.x.cflag&PSL_C)?“no”:“yes”));
apm_info.apm_version = reg.x.eax;
apm_info.apm_flags = reg.x.ecx;
DPRINTF((" BX signature=<%c%c>%s\n",
BH(reg), BL(reg),
((reg.x.ebx&0xffff)==(‘P’<<8)+‘M’)?" (ok)":" (MISMATCH)"));
DPRINTF((" BIOS APM version %x.%x\n",
(apm_info.apm_version>>8)&255, apm_info.apm_version&255 ));
if(apm_info.apm_flags & 1) { DPRINTF((" 16bit mode interface\n")); }
if(apm_info.apm_flags & 2) { DPRINTF((" 32bit mode interface\n")); }
if(apm_info.apm_flags & 4) { DPRINTF((" Slow CPU on Idle\n")); }
DPRINTF((" BIOS PM %s\n", (apm_info.apm_flags&8)?“disabled”:“enabled”));
DPRINTF((" BIOS PM %s\n",
(apm_info.apm_flags&16)?“disengaged”:“engaged”));if ((reg.x.ebx&0xffff) != (‘P’<<8)+‘M’) {
fprintf(stderr, “APM is not supported on this machine.\n”);
exit(1);
}/* disconnect first, in case somebody was connected */
memset(®, 0, sizeof(reg));
reg.x.eax = APM_DISCONNECT;
reg.x.ebx = APM_DEV_APM_BIOS;
s = int386(APM_SYSTEM_BIOS, ®, ®);
if (s) { fprintf(stderr,“error on int386\n”); exit(s); }/* connect 32 /
DPRINTF((“APM: Connect32\n”));
memset(®, 0, sizeof(reg));
reg.x.eax = APM_32BIT_CONNECT;
reg.x.ebx = APM_DEV_APM_BIOS;
s = int386(APM_SYSTEM_BIOS, ®, ®);
if (s) { fprintf(stderr,“error on int386\n”); exit(s); }
if (reg.x.cflag&PSL_C) {
fprintf(stderr, “APM: Connect32 failed (code 0x%02lX)\n”, reg.x.eax);
return 1;
}
DUMPREGS(®);
/
- AX: code32 realmode CS
- (E)BX: entry point offset
- CX: code16 realmode CS
- ESI: 16 code32 length-1
16 code16 length-1- DX: data realmode DS
- DI: data length
*/
DPRINTF((" 32Bit entry=%04X:%04X (len %04X+1)\n", reg.x.eax, reg.x.ebx,
reg.x.esi&0xffff));
DPRINTF((" 16Bit entry=%04X:%04X (len %04X+1)\n", reg.x.ecx, reg.x.ebx,
(reg.x.esi>>16)&0xffff));
DPRINTF((" Data region=%04X:( 0) (len %04X+1)\n", reg.x.edx, reg.x.edi));/* set my version (needed to access 1.1- BIOS calls) (APM 1.1) /
apm_info.apm_driver_version = 0x0101; / set to APM 1.1 /
DPRINTF((“APM: set Driver Version to %x.%x\n”,
(apm_info.apm_driver_version>>8)&255,
apm_info.apm_driver_version&255
));
memset(®, 0, sizeof(reg));
reg.x.eax = APM_DRIVER_VERSION;
reg.x.ebx = APM_DEV_APM_BIOS;
reg.x.ecx = apm_info.apm_driver_version;
s = int386(APM_SYSTEM_BIOS, ®, ®);
/
- AH.AL: APM connection version
/
if (s) { fprintf(stderr,“error on int386\n”); exit(s); }
if (reg.x.cflag&PSL_C) {
fprintf(stderr, “APM: set Driver Version failed (code 0x%02X)\n”,
(int)(reg.x.eax>>8));
/exit(1);/
} else {
DPRINTF((" Connected as version %x.%x\n", (reg.x.eax>>8)&255,
reg.x.eax&255 ));
}
/ set power status (to OFF etc) (APM 1.1) */
DPRINTF((“APM: Set Power State (All, %d)\n”, option.setstate));
memset(®, 0, sizeof(reg));
reg.x.eax = APM_SET_POWER_STATE;
reg.x.ebx = APM_DEV_APM_BIOS +0x0001;
reg.x.ecx = option.setstate;
if (!option.dryrun)
s = int386(APM_SYSTEM_BIOS, ®, ®);
if (s) { fprintf(stderr,“error on int386\n”); exit(s); }
if (reg.x.cflag&PSL_C)
fprintf(stderr, “APM: Set Power State failed (code 0x%02X)\n”,
(int)(reg.x.eax>>8));DPRINTF((“APM: Disconnect\n”));
memset(®, 0, sizeof(reg));
reg.x.eax = APM_DISCONNECT;
reg.x.ebx = APM_DEV_APM_BIOS; /APM BIOS/
s = int386(APM_SYSTEM_BIOS, ®, ®);
if (s) { fprintf(stderr,“error on int386\n”); exit(s); }
return s;
}===========================================
Thanks,
Len.
“Bill Caroselli” <> qtps@earthlink.net> > wrote in message
news:bhqinf$m90$> 2@inn.qnx.com> …If this is QNX4 there was an int386() function that will help you make
this kinf of call, I think. Check the docs.
Len Meakin <> len.meakin@cirrustesting.com> _remove> wrote:
LM > Already linked with T1 and running as root, so I guess direct int15calls
LM > are out of the question. I have found some apm code (Apm-1.3.tar.gz)
on
LM > QUICS, but this seems to only support QNX 4.25 (Proc4.25), but we
have to
LM > run with Proc4.24 because of our int10 usage. I’ll be taking a closer
look
LM > at this code today to determine if it will provide any hints on how
to
LM > accomplish this.
LM > Thank you,
LM > Len.LM > “David Gibbs” <> dagibbs@qnx.com> > wrote in message
LM > news:bhis3c$5uq$> 2@nntp.qnx.com> …Len Meakin <> len.meakin@cirrustesting.com> > wrote:
Is there any way to implement power management functionality using
QNX
LM > 4.25
(Proc32=4.24). As we require the following functionality:
Get current battery status (PC104 based system).
Complete power down of system, without user intervention (PC104 based
system).Is this possible, I have tried int15 calls but the code just seems to
sigsegv when ever it is run. Any help on this will be gratefullyLM > received.
QNX does not provide any power management functionality for QNX 4.
I don’t know if the int15 calls are callable from 32-bit protected
mode,
nor how you went about trying to call them.
At minimum, you will need to link your program with -T1 and run as
root.
-David
QNX Training Services
http://www.qnx.com/support/training/
Please followup in this newsgroup if you have further questions.
\Bill Caroselli – Q-TPS Consulting
1-(626) 824-7983
qtps@earthlink.net
Len Meakin <len.meakin@cirrustesting.com_REMOVE> wrote in message
news:bil771$ee6$1@inn.qnx.com…
I’m still really stuck with this, has anyone got any ideas on how I
could overcome this problem ? Any suggestions will be welcomed (apart
from using a hardware solution as this is not possible).
Is the APM calls into the apmbios callable from protected mode?
-Adam
Adam Mallory wrote:
Len Meakin <> len.meakin@cirrustesting.com> _REMOVE> wrote in message
news:bil771$ee6$> 1@inn.qnx.com> …I’m still really stuck with this, has anyone got any ideas on how I
could overcome this problem ? Any suggestions will be welcomed (apart
from using a hardware solution as this is not possible).
Is the APM calls into the apmbios callable from protected mode?-Adam
It does not seem so (I have now found several old threads attempting
this with the same results - sigsegv). So the direct calls are out of
the question it would seem. However I have now changed to using Proc32
v4.25O and the Apm server from quics. This does give me the ability to
completley power down the PC and get battery status. However when i
start the Apm server, I get the following message:
found pm … scanning queue1 and queue2
This is displayed continuasly (without let up) on what ever console is
currently active. I have already tried redirecting the Apm servers
output to /dev/null (1>/dev/null 2>&1) but without any luck. I have also
grepped the Apm server source code and was unable to find this message.
So my question is what is printing this message and how the hell can i
remove it.
And just one further question, I tried re-compiling the Apm source code
but my proc_msg.h does not contain a referance to the following structs:
extern struct _proc_pm_control
extern struct _proc_apmbios_data
Maybe more … ?
Thank you for your help.
Len.
Len Meakin <len.meakin@cirrustesting.com_REMOVE> wrote in message
news:bin2np$q0o$1@inn.qnx.com…
However when i
start the Apm server, I get the following message:found pm … scanning queue1 and queue2
This is displayed continuasly (without let up) on what ever console is
currently active. I have already tried redirecting the Apm servers
output to /dev/null (1>/dev/null 2>&1) but without any luck. I have also
grepped the Apm server source code and was unable to find this message.
So my question is what is printing this message and how the hell can i
remove it.
That is a debugging statement from Proc, which was left there some time ago.
There isn’t much you can do to stop it… you could perhaps set the -o
option to Proc32 to specify that output should go to an unsed serial port.
And just one further question, I tried re-compiling the Apm source code
but my proc_msg.h does not contain a referance to the following structs:extern struct _proc_pm_control
extern struct _proc_apmbios_data
Maybe more … ?
Strange - ok, contact support and they can get you an updated version of
that header to rebuild the Apm source.
-Adam
Adam Mallory wrote:
Len Meakin <> len.meakin@cirrustesting.com> _REMOVE> wrote in message
news:bin2np$q0o$> 1@inn.qnx.com> …
snipHowever when i
start the Apm server, I get the following message:found pm … scanning queue1 and queue2
This is displayed continuasly (without let up) on what ever console is
currently active. I have already tried redirecting the Apm servers
output to /dev/null (1>/dev/null 2>&1) but without any luck. I have also
grepped the Apm server source code and was unable to find this message.
So my question is what is printing this message and how the hell can i
remove it.
That is a debugging statement from Proc, which was left there some time ago.
There isn’t much you can do to stop it… you could perhaps set the -o
option to Proc32 to specify that output should go to an unsed serial port.
And just one further question, I tried re-compiling the Apm source code
but my proc_msg.h does not contain a referance to the following structs:extern struct _proc_pm_control
extern struct _proc_apmbios_data
Maybe more … ?
Strange - ok, contact support and they can get you an updated version of
that header to rebuild the Apm source.-Adam
Thanks for the response Adam, I have two serial ports both of which are
in use so redirecting Proc32’s output to a serial port is not an option
available to me. I know this may seem a really strange question but why
was this debug message left in, and would it be possible to only display
it if Proc32 is set to be in one of the verbose modes. As this message
is constantly bombarded to the screen it is not possible to do anything
once the apm server is started. I just seem to keep hitting brick walls
with this apm stuff =((.
I’ll contact support with regard this problem and the updated header.
Thank you for all your help.
Len.
Len Meakin <len.meakin@cirrustesting.com_REMOVE> wrote in message
news:biup12$bal$1@inn.qnx.com…
Thanks for the response Adam, I have two serial ports both of which are
in use so redirecting Proc32’s output to a serial port is not an option
available to me. I know this may seem a really strange question but why
was this debug message left in, and would it be possible to only display
it if Proc32 is set to be in one of the verbose modes. As this message
is constantly bombarded to the screen it is not possible to do anything
once the apm server is started. I just seem to keep hitting brick walls
with this apm stuff =((.
I’ll contact support with regard this problem and the updated header.
Thank you for all your help.
Sorry there isn’t a way to take it out (short of patching the call to a NOP)
with the Proc you have. 
The message was left in only by oversight at the time (years back), sorry
about that. One thing you can try is to patch out the call to the kprintf
in the Proc binary. The location in Proc32 version O is offset 0x675F, the
pattern you want is E8 E2 46 00 00 - this is the call to display the
message. You can patch each of those values with NOP opcodes (NOP opcode is
0x90) so that the pattern looks like 90 90 90 90 90.
Of course I would recommend making a copy of the Proc binary before doing
this FYI - this modification isn’t supported by QSSL.
-Adam
Adam Mallory wrote:
Len Meakin <> len.meakin@cirrustesting.com> _REMOVE> wrote in message
news:biup12$bal$> 1@inn.qnx.com> …
Thanks for the response Adam, I have two serial ports both of which are
in use so redirecting Proc32’s output to a serial port is not an option
available to me. I know this may seem a really strange question but why
was this debug message left in, and would it be possible to only display
it if Proc32 is set to be in one of the verbose modes. As this message
is constantly bombarded to the screen it is not possible to do anything
once the apm server is started. I just seem to keep hitting brick walls
with this apm stuff =((.
I’ll contact support with regard this problem and the updated header.
Thank you for all your help.
Sorry there isn’t a way to take it out (short of patching the call to a NOP)
with the Proc you have. >The message was left in only by oversight at the time (years back), sorry
about that. One thing you can try is to patch out the call to the kprintf
in the Proc binary. The location in Proc32 version O is offset 0x675F, the
pattern you want is E8 E2 46 00 00 - this is the call to display the
message. You can patch each of those values with NOP opcodes (NOP opcode is
0x90) so that the pattern looks like 90 90 90 90 90.Of course I would recommend making a copy of the Proc binary before doing
this >-Adam
Thank you for your response this works great. Hopefully this debug
message will be removed if there are further releases of Proc32. Once
again thanks to everyone for your help.
Len.
Len Meakin <len.meakin@cirrustesting.com_REMOVE> wrote in message
news:bj433f$8m2$1@inn.qnx.com…
Thank you for your response this works great. Hopefully this debug
message will be removed if there are further releases of Proc32. Once
again thanks to everyone for your help.
NP. Already done Thanks for pointing it out!
-Adam