Re: Can NMI drop a hanging FreeBSD kernel into DDB?
Warner Losh wrote: In message [EMAIL PROTECTED] Sergey Babkin writes: : Seems like most of the modern machines just don't have that : pin on the PCI bus connected anywhere. But on most of them : (though not all) the pin on ISA works. Some high-end machines : like Unisys or Compaq have an NMI button on the box (sometimes : under the cover). IOCHK* isn't on the PCI bus at all. You have to do weird things for it to generate an NMI that I've never quite worked out. I sure wish I Probably SERR# is used instead. We have small cards at work with a big button, ISA connector on one side and PCI connector on the other one. I'm not sure exactly to which signal on the bus is it connected. But the PCI side commonly don't work for the modern machines. could get the pcccard bus (and/or cardbus) to genereate NMIs for laptop hacking at times. If it's derived from PCI you can try to use SERR# or its analog. -SB To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: Can NMI drop a hanging FreeBSD kernel into DDB?
Peter Jeremy wrote: On 2000-May-11 07:10:27 +1000, Wilko Bulte [EMAIL PROTECTED] wrote: And if you force the IOCHK* line on an AT slot to GND? Would that work on modern PCI machines? Grounding IOCHK* does cause an NMI on the only PCI machine I've tried it on. It looks like this is maskable in the Intel PIIX3 (and presumably later) chips, so it may depend on the BIOS. Seems like most of the modern machines just don't have that pin on the PCI bus connected anywhere. But on most of them (though not all) the pin on ISA works. Some high-end machines like Unisys or Compaq have an NMI button on the box (sometimes under the cover). I don't know whether this will still work if the PCI bus is hung. I Most probably it won't. The keyboard is connected on most machines below PCI, so if PCI hangs you won't be able to do anything anyway. The same problem would be in the other direction, for PCI video card. -SB To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: Can NMI drop a hanging FreeBSD kernel into DDB?
In message [EMAIL PROTECTED] Peter Jeremy writes: : I thought the PCI bus was isolated from the CPU-memory bus so this : couldn't occur. The trap handler should start OK, but just can't do : any I/O because the PCI bus is hung. I'm not sure why things happen this way. I just know that I've not been able to get a NMI to break into the debugger when certain hardware that I've worked on in the past when it desided to hang. I don't know if that's because this also hung the host bridge so that no interrupt code could fetch it from memory, or what the deal was. Warner To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: Can NMI drop a hanging FreeBSD kernel into DDB?
On Thu, May 11, 2000 at 07:01:43AM +1000, Peter Jeremy wrote: On Fri, 05 May 2000 22:56:42 -0600, Warner Losh [EMAIL PROTECTED] wrote: In message [EMAIL PROTECTED] Warner Losh writes: : However, it won't work if you are hacking pci hardware and manage to : hang the PCI bus. Unless, of course, the trap handler is in cache as well as the debugger routines you need. I thought the PCI bus was isolated from the CPU-memory bus so this couldn't occur. The trap handler should start OK, but just can't do any I/O because the PCI bus is hung. And if you force the IOCHK* line on an AT slot to GND? Would that work on modern PCI machines? I used to do this on older EISA boxes. -- Wilko Bulte FreeBSD, the power to serve http://www.freebsd.org http://www.tcja.nl To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: Can NMI drop a hanging FreeBSD kernel into DDB?
On Fri, 05 May 2000 22:56:42 -0600, Warner Losh [EMAIL PROTECTED] wrote: In message [EMAIL PROTECTED] Warner Losh writes: : However, it won't work if you are hacking pci hardware and manage to : hang the PCI bus. Unless, of course, the trap handler is in cache as well as the debugger routines you need. I thought the PCI bus was isolated from the CPU-memory bus so this couldn't occur. The trap handler should start OK, but just can't do any I/O because the PCI bus is hung. There's not a good way to force it either since if you knew you were about to hang the pci bus, you'd not do the action that would hang it :-). At least on the 486, it is possible to freeze data in the internal cache by loading the cache (either via the test registers or by judicious memory accesses) and then disabling it (CR0.CD = 1). It's not possible to freeze specific cache lines. I presume similar facilities are available in the later processors and for external caches. Peter To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: Can NMI drop a hanging FreeBSD kernel into DDB?
In message [EMAIL PROTECTED] Sergey Babkin writes: : Seems like most of the modern machines just don't have that : pin on the PCI bus connected anywhere. But on most of them : (though not all) the pin on ISA works. Some high-end machines : like Unisys or Compaq have an NMI button on the box (sometimes : under the cover). IOCHK* isn't on the PCI bus at all. You have to do weird things for it to generate an NMI that I've never quite worked out. I sure wish I could get the pcccard bus (and/or cardbus) to genereate NMIs for laptop hacking at times. Warner To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: Can NMI drop a hanging FreeBSD kernel into DDB?
On Fri, 5 May 2000 12:20:29 -0700, Alfred Perlstein [EMAIL PROTECTED] wrote:
How can I make a hanging kernel into DDB? Will grounding the NMI do it?
That's a bit extreme.
hit ctrl+alt+esc on the console, or send a serial break if using
a serial console, make sure you have BREAK_TO_DEBUGGER if you're
using a serial console.
Assuming that the kernel is still processing interrupts.
If your driver has managed something like "di(); while (1) {}"
or maybe "splhigh(); while (1) {}", then you need an NMI to get
you out of the loop so the console interrupt is seen.
Peter
To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-hackers" in the body of the message
Re: Can NMI drop a hanging FreeBSD kernel into DDB?
On 2000-May-11 07:10:27 +1000, Wilko Bulte [EMAIL PROTECTED] wrote: And if you force the IOCHK* line on an AT slot to GND? Would that work on modern PCI machines? Grounding IOCHK* does cause an NMI on the only PCI machine I've tried it on. It looks like this is maskable in the Intel PIIX3 (and presumably later) chips, so it may depend on the BIOS. I don't know whether this will still work if the PCI bus is hung. I know it doesn't work when I wedge that machine, but I'm not sure what is wedging (since I'm not playing with untested PCI hardware, I think it's not the PCI bus). Peter To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: Can NMI drop a hanging FreeBSD kernel into DDB?
Warner Losh wrote: In message [EMAIL PROTECTED] Warner Losh writes: : However, it won't work if you are hacking pci hardware and manage to : hang the PCI bus. Unless, of course, the trap handler is in cache as well as the debugger routines you need. On rare occasions this is true, but almsot all the time it isn't. There's not a good way to force it either since if you knew you were about to hang the pci bus, you'd not do the action that would hang it :-). No cache-line-locking on x86? Sigh. -- "Where am I, and what am I doing in this handbasket?" Wes Peters Softweyr LLC [EMAIL PROTECTED] http://softweyr.com/ To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: Can NMI drop a hanging FreeBSD kernel into DDB?
* Jin Guojun [EMAIL PROTECTED] [000505 12:03] wrote: Hi, I am currently experimenting kernel hanging problem even through I have compiled the kernel with DDB enabled. How can I make a hanging kernel into DDB? Will grounding the NMI do it? Thanks for any suggestions, That's a bit extreme. hit ctrl+alt+esc on the console, or send a serial break if using a serial console, make sure you have BREAK_TO_DEBUGGER if you're using a serial console. -- -Alfred Perlstein - [[EMAIL PROTECTED]|[EMAIL PROTECTED]] "I have the heart of a child; I keep it in a jar on my desk." To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: Can NMI drop a hanging FreeBSD kernel into DDB?
* Alfred Perlstein [EMAIL PROTECTED] [000505 12:21] wrote: * Jin Guojun [EMAIL PROTECTED] [000505 12:03] wrote: Hi, I am currently experimenting kernel hanging problem even through I have compiled the kernel with DDB enabled. How can I make a hanging kernel into DDB? Will grounding the NMI do it? Thanks for any suggestions, That's a bit extreme. hit ctrl+alt+esc on the console, or send a serial break if using a serial console, make sure you have BREAK_TO_DEBUGGER if you're using a serial console. Forgot to mention that an NMI will also work. -Alfred To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: Can NMI drop a hanging FreeBSD kernel into DDB?
In message [EMAIL PROTECTED] Alfred Perlstein writes: : hit ctrl+alt+esc on the console, or send a serial break if using : a serial console, make sure you have BREAK_TO_DEBUGGER if you're : using a serial console. Sometimes this isn't enough. I use the NMI trick often. However, it won't work if you are hacking pci hardware and manage to hang the PCI bus. Warner To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: Can NMI drop a hanging FreeBSD kernel into DDB?
In message [EMAIL PROTECTED] Jin Guojun writes: : In a normal operation, both work well. When system hangs, neither of them : works. It looks that system is in some wiered status. : : Does any one happen to know how to trace this kind of problem? PCI bus analizer is one way. Another is to have an ICE that will tell you the last value of the progam counter. Without some sort of hardware assist, these thing can be very hard to track down. Warner To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: Can NMI drop a hanging FreeBSD kernel into DDB?
In message [EMAIL PROTECTED] Warner Losh writes: : However, it won't work if you are hacking pci hardware and manage to : hang the PCI bus. Unless, of course, the trap handler is in cache as well as the debugger routines you need. On rare occasions this is true, but almsot all the time it isn't. There's not a good way to force it either since if you knew you were about to hang the pci bus, you'd not do the action that would hang it :-). Warner To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
