January 2007 - Darwin-kernel

Re: Understanding cores...
by Andreas Fink 08 Jan '07

08 Jan '07

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

1 0

Re: User-space to kernel communication
by Josh Graessley 08 Jan '07

08 Jan '07

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com -josh On Jan 7, 2007, at 6:40 PM, Greg wrote: Thanks, _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-kernel mailing list (Darwin-kernel(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-kernel/site_archiver%40lists.… This email sent to site_archiver(a)lists.apple.com This depends on what you are trying to communicate with in the kernel. If you are already using IOKit, I would imagine an IOUserClient would be the way to go. If you're writing a non-IOKit kext, you may want to look in to a kernel control socket (sys/ kern_control.h). Hi, what is the best way(s) to send variable length data (around 50 bytes or so) to a kernel extension from a user-land application? Are there different ways to do this if the kext uses the IOUserClient schema to talk to user-land apps? If you could also point me to documentation or a tutorial on how to do this I would be much obliged! - Greg _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-kernel mailing list (Darwin-kernel(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-kernel/jgraessley% 40apple.com This email sent to jgraessley(a)apple.com smime.p7s

1 0

Re: Understanding cores...
by Derek Kumar 08 Jan '07

08 Jan '07

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-kernel mailing list (Darwin-kernel(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-kernel/site_archiver%40lists.… (gdb) paniclog panic(cpu 2 caller 0x001A3135): Unresolved kernel trap (CPU 2, Type 14=page fault), registers: CR0: 0x8001003b, CR2: 0xffffffe0, CR3: 0x011b0000, CR4: 0x000006e0 EAX: 0x00000000, EBX: 0x352c4038, ECX: 0x352c4000, EDX: 0x352dd008 CR2: 0xffffffe0, EBP: 0x00000000, ESI: 0x352f4010, EDI: 0x004ae85c EFL: 0x00010046, EIP: 0x00135f3b, CS: 0x00000008, DS: 0x03820010 Backtrace, Format - Frame : Return Address (4 potential args on stack) 0x252a3d68 : 0x128d1f (0x3c9540 0x252a3d8c 0x131df4 0x0) 0x252a3da8 : 0x1a3135 (0x3cf1f4 0x2 0xe 0x3cea24) 0x252a3eb8 : 0x19a8d4 (0x252a3ed0 0x352f4000 0x252a3f28 0x352f4000) Backtrace terminated-invalid frame pointer 0x0 Kernel version: Darwin Kernel Version 8.8.1: Mon Sep 25 19:42:00 PDT 2006; root:xnu-792.13.8.obj~1/RELEASE_I386 task vm_map ipc_space #acts pid proc command 0x03808da0 0x013e3f3c 0x037d2ef0 54 0 0x004d2200 kernel_task activation thread pri state wait_queue wait_event 0x03822e68 0x03822e68 0 IR reserved_stack=0x24ff8000 kernel_stack=0x252a0000 stacktop=0x252a3d68 0x252a3d68 0x128d1f <panic+382> 0x252a3da8 0x1a3135 <kernel_trap+1538> 0x252a3eb8 0x19a8d4 <trap_from_kernel+19> stackbottom=0x252a3eb8 As Brian noted, this is a page fault that could not be satisfied, and on the kernel version noted in the panic log, the faulting instruction corresponds to the idle_thread() function. This is the machine independent entry to the processor idle code, which essentially looks for threads to dispatch and invokes the machine specific power management state machine/idle power state as appropriate. The fault was triggered whilst trying to load a local (using the base+displacement mode), but the framepointer (EBP) was 0, leading to the fault. Nothing directly points to your NKE as far as I can tell, but one possibility is memory corruption leading to the zeroing of the EBP record in the register state stored at the base of the kernel stack (which is reloaded at context switch time). You say your system "crashes a lot"--does it crash consistently in this manner, and only when your NKE is loaded? Derek gdb) x/i 0x135f3b 0x135f3b <idle_thread+120>: mov -32(%ebp),%edx which corresponds to 2544 while ( (*threadp == THREAD_NULL) && 2545 (*gcount == 0) && (*lcount == 0) ) and "threadp" is a local, and EBP, the framepointer, happens to be 0, leading to a page fault This email sent to site_archiver(a)lists.apple.com

1 0

Re: Understanding cores...
by Brian Bechtel 08 Jan '07

08 Jan '07

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com Domainkey-signature: a=rsa-sha1; q=dns; c=nofws; s=beta; d=gmail.com; h=received:message-id:date:from:to:subject:cc:in-reply-to:mime-version:content-type:content-transfer-encoding:content-disposition:references; b=SrsBBsYKXPtB29SIYOZSZVt4v+FzcArzOtOKRYh2zkld64fjZpFUEqQYzj9d4A64icmtjL004k24ympRVXrTlo97NDc7fwtjmutemD7o4g6sweUHhjlRXTeO9QicnLFJmgt0xP11ci5Z4xQc9cwq01PUYIyuQsXv4Qb1D1LqRh8= I'm looking at a core of a MacPro. It crashes a lot and it use an NKE. So I would like to understand why the system is crashing, if it is related to the NKE and finally how to fix it... [...] (gdb) paniclog panic(cpu 2 caller 0x001A3135): Unresolved kernel trap (CPU 2, Type 14=page fault), registers: CR0: 0x8001003b, CR2: 0xffffffe0, CR3: 0x011b0000, CR4: 0x000006e0 EAX: 0x00000000, EBX: 0x352c4038, ECX: 0x352c4000, EDX: 0x352dd008 CR2: 0xffffffe0, EBP: 0x00000000, ESI: 0x352f4010, EDI: 0x004ae85c EFL: 0x00010046, EIP: 0x00135f3b, CS: 0x00000008, DS: 0x03820010 Backtrace, Format - Frame : Return Address (4 potential args on stack) 0x252a3d68 : 0x128d1f (0x3c9540 0x252a3d8c 0x131df4 0x0) 0x252a3da8 : 0x1a3135 (0x3cf1f4 0x2 0xe 0x3cea24) 0x252a3eb8 : 0x19a8d4 (0x252a3ed0 0x352f4000 0x252a3f28 0x352f4000) Backtrace terminated-invalid frame pointer 0x0 You have a page fault. The offending address is in CR2: 0xffffffe0 The IP at the time of the fault was in EIP: 0x00135f3b The "showcurrentstacks" demonstrates that the other 3 cores were idle. There isn't enough information here to diagnose this further. The systems runs MacOS X 10.4.8 Server. So could it be that the system locked up and the Watchdog forced it to reboot. Would we then see a trace like the one above? No, if the watchdog forced it to reboot, you wouldn't be able to attach and get the "showcurrentstacks". A reboot is not a panic, although a panic can cause a reboot. You need to examine the code around that IP and see why it's faulting. On a Mac Pro I have access to, that's in the code "idle_thread" which tells me... not much. task vm_map ipc_space #acts pid proc command 0x03808da0 0x013e3f3c 0x037d2ef0 54 0 0x004d2200 kernel_task activation thread pri state wait_queue wait_event 0x03822e68 0x03822e68 0 IR reserved_stack=0x24ff8000 kernel_stack=0x252a0000 stacktop=0x252a3d68 0x252a3d68 0x128d1f <panic+382> 0x252a3da8 0x1a3135 <kernel_trap+1538> 0x252a3eb8 0x19a8d4 <trap_from_kernel+19> stackbottom=0x252a3eb8 try "switchtoact 0x03822e68" and examine the state of the thread. Sorry, you'll just have to debug further. There isn't enough information for anyone to help you more than that. _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-kernel mailing list (Darwin-kernel(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-kernel/site_archiver%40lists.… On 1/6/07, Michael Tuexen <Michael.Tuexen(a)lurchi.franken.de> wrote: This email sent to site_archiver(a)lists.apple.com

1 0

Delay Caused By Software Interrupt
by Satyam Lad 08 Jan '07

08 Jan '07

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com Domainkey-signature: a=rsa-sha1; q=dns; c=nofws; s=beta; d=gmail.com; h=received:message-id:date:from:to:subject:mime-version:content-type; b=FR37O8s1iiLJ4pfe9EpyNMpa+mtU3/8Kjdz3+122454lC2HDKe0Eu1UR8ZqFu3ulaVQ+wSb4wKJlxf30dmpvE8IyG+V6XbHQKVkLizRUhyCjwoAV8z3Q23a9OGWdF1cpihuhT0APsnOCCmk+iZIoowYLkoRjhfmduy9Zxmiyxy4= _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-kernel mailing list (Darwin-kernel(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-kernel/site_archiver%40lists.… This email sent to site_archiver(a)lists.apple.com

1 0

User-space to kernel communication
by Greg 08 Jan '07

08 Jan '07

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com Thanks, - Greg _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-kernel mailing list (Darwin-kernel(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-kernel/site_archiver%40lists.… Hi, what is the best way(s) to send variable length data (around 50 bytes or so) to a kernel extension from a user-land application? Are there different ways to do this if the kext uses the IOUserClient schema to talk to user-land apps? If you could also point me to documentation or a tutorial on how to do this I would be much obliged! This email sent to site_archiver(a)lists.apple.com

1 0

Understanding cores...
by Michael Tuexen 06 Jan '07

06 Jan '07

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com Dear all, Trying to find what was running on the system results in (gdb) showcurrentstacks task vm_map ipc_space #acts pid proc command 0x03808da0 0x013e3f3c 0x037d2ef0 54 0 0x004d2200 kernel_task activation thread pri state wait_queue wait_event 0x03827c64 0x03827c64 0 IR reserved_stack=0x25008000 kernel_stack=0x25258000 stacktop=0x2525bf18 0x2525bf18 0x1a42f5 <machine_idle_cstate+32> 0x2525bf38 0x19d871 <machine_idle+128> 0x2525bf58 0x135f23 <idle_thread+96> 0x2525bfc8 0x19a74c <call_continuation+28> stackbottom=0x2525bfc8 task vm_map ipc_space #acts pid proc command 0x03808da0 0x013e3f3c 0x037d2ef0 54 0 0x004d2200 kernel_task activation thread pri state wait_queue wait_event 0x03822b04 0x03822b04 0 IR reserved_stack=0x24fd8000 kernel_stack=0x25378000 stacktop=0x2537bf18 0x2537bf18 0x1a42f5 <machine_idle_cstate+32> 0x2537bf38 0x19d871 <machine_idle+128> 0x2537bf58 0x135f23 <idle_thread+96> 0x2537bfc8 0x19a74c <call_continuation+28> stackbottom=0x2537bfc8 task vm_map ipc_space #acts pid proc command 0x03808da0 0x013e3f3c 0x037d2ef0 54 0 0x004d2200 kernel_task activation thread pri state wait_queue wait_event 0x03822e68 0x03822e68 0 IR reserved_stack=0x24ff8000 kernel_stack=0x252a0000 stacktop=0x252a3d68 0x252a3d68 0x128d1f <panic+382> 0x252a3da8 0x1a3135 <kernel_trap+1538> 0x252a3eb8 0x19a8d4 <trap_from_kernel+19> stackbottom=0x252a3eb8 task vm_map ipc_space #acts pid proc command 0x03808da0 0x013e3f3c 0x037d2ef0 54 0 0x004d2200 kernel_task activation thread pri state wait_queue wait_event 0x03825050 0x03825050 0 IR reserved_stack=0x24fe0000 kernel_stack=0x252c0000 stacktop=0x252c3f18 0x252c3f18 0x1a42f5 <machine_idle_cstate+32> 0x252c3f38 0x19d871 <machine_idle+128> 0x252c3f58 0x135f23 <idle_thread+96> 0x252c3fc8 0x19a74c <call_continuation+28> stackbottom=0x252c3fc8 So how can I figure out why the system crashed? Thank you very much for your help in advance. Best regards Michael _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-kernel mailing list (Darwin-kernel(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-kernel/site_archiver%40lists.… I'm looking at a core of a MacPro. It crashes a lot and it use an NKE. So I would like to understand why the system is crashing, if it is related to the NKE and finally how to fix it... The problem is that all data structures of the NKE look fine, from a network trace which was also made at the time of the crash we know that there was no incoming SCTP packet (this is the transport layer which is implemented by the NKE), no system call by the user and no timer was running of. The paniclog says: (gdb) source KernelDebugKit/kgmacros Loading Kernel GDB Macros package. Type "help kgm" for more info. (gdb) paniclog panic(cpu 2 caller 0x001A3135): Unresolved kernel trap (CPU 2, Type 14=page fault), registers: CR0: 0x8001003b, CR2: 0xffffffe0, CR3: 0x011b0000, CR4: 0x000006e0 EAX: 0x00000000, EBX: 0x352c4038, ECX: 0x352c4000, EDX: 0x352dd008 CR2: 0xffffffe0, EBP: 0x00000000, ESI: 0x352f4010, EDI: 0x004ae85c EFL: 0x00010046, EIP: 0x00135f3b, CS: 0x00000008, DS: 0x03820010 Backtrace, Format - Frame : Return Address (4 potential args on stack) 0x252a3d68 : 0x128d1f (0x3c9540 0x252a3d8c 0x131df4 0x0) 0x252a3da8 : 0x1a3135 (0x3cf1f4 0x2 0xe 0x3cea24) 0x252a3eb8 : 0x19a8d4 (0x252a3ed0 0x352f4000 0x252a3f28 0x352f4000) Backtrace terminated-invalid frame pointer 0x0 Kernel version: Darwin Kernel Version 8.8.1: Mon Sep 25 19:42:00 PDT 2006; root:xnu-792.13.8.obj~1/RELEASE_I386 The systems runs MacOS X 10.4.8 Server. So could it be that the system locked up and the Watchdog forced it to reboot. Would we then see a trace like the one above? This email sent to site_archiver(a)lists.apple.com

1 0

Re: Kext suspension question
by Terry Lambert 06 Jan '07

06 Jan '07

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com -- Terry _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-kernel mailing list (Darwin-kernel(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-kernel/site_archiver%40lists.… On Jan 5, 2007, at 7:02 PM, Greg wrote: Hi, I'm writing a kext that uses the kauth facilities and was wondering whether it is possible to wait for user input from a userland application connected to the kext through an IOUserClient object in the middle of making the decision of whether or not a file system operation will be allowed. Essentially I would like to send a notification to the user in my kauth_listener function, wait for their decision, and then decide based on that whether to allow or disallow the file system operation in question. I haven't worked in the kernel before and am worried that this behavior may cause the entire system to hang. It will likely cause the system to hang, if you are asking for authentication on every authorization request, rather than e.g. on a specific file. Even then, it's mostly, that's not a good idea. In general, if your user space client application ends up making any FS calls, it will end up blocked behind itself authorizing itself to make those calls (e.g. if it needs to access a password file, or needs to open a resource file to display a localized string in a dialog, and so on). A good model to follow in this regard is the ssh-agent model. In that model, the ssh-agent will turn around authorization requests immediately, and if you want a session authorized to it, then you use ssh-add. This divorces the user input for granting of authorization from the daemon that turns around the requests (all requests are denied unless you are authorized). This works best if the application itself knows about authorization (i.e. if ssh-agent is not running, or if you have not authorized yourself to it via ssh-add, it'll ask you for the password instead when it can't get it from the agent). This email sent to site_archiver(a)lists.apple.com

1 0

Kext suspension question
by Greg 06 Jan '07

06 Jan '07

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com Thanks for any help, - Greg _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-kernel mailing list (Darwin-kernel(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-kernel/site_archiver%40lists.… Hi, I'm writing a kext that uses the kauth facilities and was wondering whether it is possible to wait for user input from a userland application connected to the kext through an IOUserClient object in the middle of making the decision of whether or not a file system operation will be allowed. Essentially I would like to send a notification to the user in my kauth_listener function, wait for their decision, and then decide based on that whether to allow or disallow the file system operation in question. I haven't worked in the kernel before and am worried that this behavior may cause the entire system to hang. This email sent to site_archiver(a)lists.apple.com

1 0

Re: TSO / LSO
by Andrew Gallatin 04 Jan '07

04 Jan '07

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com Terry Lambert writes: > On Jan 3, 2007, at 10:28 AM, Adi Masputra wrote: > > On Jan 3, 2007, at 6:44 AM, Andrew Gallatin wrote: > > > >> There are 3 problems with the above: > >> > >> 1) The NIC does not advertise any sort of MTU. The stack just sends > >> down the biggest frame it can (ip_len is 16 bits, so the max is 64K-1 > >> for IPv4..). The NIC is responsible for splitting this into as many > >> packets as required. > > > Adi covered most of this below, but the real concern I had here was > the NIC advertising the largest buffer that it was willing to take at > a time. To me, this is the effective MTU to the card; you said 64K, > but from my reading on TSO, you might wish to limit this due to on > card memory buffer constraints vs. the number of active streams you > expect to deal with simultaneously (i.e. host quench based on > knowledge of card transmit buffer size). The limitation is not needed, but there is a precedent for it in the NDIS stack. When we did our windows driver, we were a bit puzzled by the purpose for it :) > Minimally, there would be a need for a global override for protecting > legacy filters from TSO (i.e. a way to turn it off, if filters were > present). Again, I don't dispute any problems with filters. I know nothing about the filter architecture, and defer to you guys. > >> 3) There is no need to buffer up an entire window on the NIC. All > >> you > >> really need to buffer on the NIC is one frame's worth of data (so > >> that > >> you can properly send the frame on the wire). > > Well, you also need the header template in hand for each in progress > connection for which the data hasn't already been sent (minimally), > etc.. Sure. This adds a max of what, 128 bytes? I was afraid you pick this out, and should have mentioned it.. > This assumes that either the transmit buffer data is contiguous (a bad > assumption, unless we are packing packet headers at the end of a page, > and only mapping page aligned data following the headers), or that the > NIC supports scatter/gather DMA. How many nics made in the last decade have not supported s/g? :) > If you consider that 100% bus-on for a NIC on a 64 bit wide PCI bus, > you are talking about monopolizing the bus with data to keep a card > properly fed. Breaking this up with scatter/gather would make things > worse by increasing the transaction overhead, rather than amortizing > it over a large buffer. If you're going to saturate the link for 10GbE, you're going to push the IO bus hard. Even on a MacPro, we only see in the neighborhood of 11.25Gb/s PCIe DMA read bandwidth for our 8 lane NIC. If you think S/G is bad, consider 64K (well 654240 to make the numbers round) sent on a 1460MSS link. Without TSO, you end up typically doing 44 54 byte DMAs, and 44 1460 byte DMAs. With TSO, you end up doing a single 54 byte DMA, and 15 4KB DMAs (assuming 4KB mbufs), plus a 2800 byte DMA at the end.. So TSO has reduced the number of DMAs to 17 vs 88. That's ammortizing.. > Practically, I think the card has to have on-board buffers, not a > single frame buffer, for this to be a happy thing; absolutely worst Sure. For example, we have 32KB. This handles a 64KB offload nicely. An MTU sized buffer is just an example of what you could do if you were cheap. In order to keep a pipeline fed, you need to have several times the MTU. You don't need to buffer the entire "fake" jumbogram before you carve it and send it, and you certainly don't need an entire TCP window. This is a stateless offload.. the NIC does not even know what the TCP window is! It has no reason to care about it. > case, you have a .1% lossy link, and you are going back and redoing > the DMA to do a retransmit otherwise, and having to step through and > recreate the packets to the transmit packet (consider additional > fragging on intermediate routes and RED queueing or QOS-based drops in > intermediate routers may adjust the frame base on byte boundaries, > rather than nice clean packet boundaries - ACK's in the middle of what > you thought as a packet, etc.). > > I expect that for TSO, you'd just eat the retransmit overhead, but you > might want to have the stack know that it was happening, and cut down > on the replicated data that was getting sent (and the amount of bus > traffic, as a result). Again, the nic has no knowledge of any drops or packet loss. As we discussed before, this is a *stateless* offload. All TCP state is mainted by the host TCP/IP stack. If there are any drops, the host will notice it from the ack, sacks, or lack thereof. > In my experience, there's usually a much larger distance between > initial implementation and what you can commercially deploy than you > might think. Our commercially deployed 10GbE NIC with TSO (and LRO) has lots of satisfied customers using OSes like Linux and Windows which support TSO. BTW, I need to do LRO for MacOSX. This nice thing about LRO is that you can do it in the NIC and driver on nearly any OS that doesn't have stupid restrictions on buffer chaining (like NDIS's restrictions). Unfortunately, the MacOSX receive performance is already OK, its the send performance which stinks. Drew _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-kernel mailing list (Darwin-kernel(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-kernel/site_archiver%40lists.… This email sent to site_archiver(a)lists.apple.com

1 0