October 2012 - Darwin-kernel

Returning ENOMEM in sysctl handler
by Alexander Dimitrov 22 Oct '12

22 Oct '12

Hello to all, I have tried sending/getting some data to/from a bsd kernel extension using a sysctl, but when I return ENOMEM from the handler the user space got success from the sysctl(3) call. I couldn't find explanation for this case in the man page, so I made a very simple kext which registers a systl using sysctl_register_oid under _debug, which handler always returns ENOMEM. When calling this sysctl from user space I observed the following things: [1] If I call sysctlbyname with oldp set to NULL then I get an error code -1 and errno is set to ENOMEM, but [2] if I call sysctlbyname with valid oldp, in my case pointing to a stack buffer, the call to sysctlbyname returns 0. According to sysctl(3) I expect in case [2] to get ENOMEM again - "If the amount of data available is greater than the size of the buffer supplied, the call supplies as much data as fits in the buffer provided and returns with the error code ENOMEM." The system I'm testing on is 10.7.4 with Xcode 4.4.1 installed. I could also provide the source of the kext and the user space program calling the sysctl if anyone wants to test this. I tried to look through the xnu source but didn't found any place where ENOMEM is handled explicitly and changed to 0. I also debugged the command line after calling sysctl(3) and saw that till the call to syscall the execution path is the same, but if the sysctl handler returns ENOMEM the syscall returns with CF cleared and the the whole call returns success, but if the handler returns ENOATTR, then syscall returns with CF and we go on the fail path (errno is set to ENOATTR, etc). Can somebody point me whether this is well-known behavior of sysctl, is it undocumented or I'm misreading the man page? Thank you, Alex. _______________________________________________ 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: https://lists.apple.com/mailman/options/darwin-kernel/site_archiver%40lists… This email sent to site_archiver(a)lists.apple.com

1 0

Re: Injecting file to current process from KEXT
by Quinn "The Eskimo!" 22 Oct '12

22 Oct '12

On 19 Oct 2012, at 10:14, Jakub Bednar <jakub.bednar(a)avg.com> wrote: > Do you know about any way, how to open a file from KEXT so it gets inserted into a file descriptor table of the current process? There's no supported way to do this; the routines necessary to manipulate the file descriptor tables for a process are not exported to KEXTs. Share and Enjoy -- Quinn "The Eskimo!" <http://www.apple.com/developer/> Apple Developer Relations, Developer Technical Support, Core OS/Hardware _______________________________________________ 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: https://lists.apple.com/mailman/options/darwin-kernel/site_archiver%40lists… This email sent to site_archiver(a)lists.apple.com

1 0

Injecting file to current process from KEXT
by Jakub Bednar 19 Oct '12

19 Oct '12

Hi everyone, Do you know about any way, how to open a file from KEXT so it gets inserted into a file descriptor table of the current process? I am trying to save an unneeded context switch caused by the process calling open() right after it returns from the KEXT. In Kernel.framework, there is a sys/vnode_if.h file mentioning a VNOP_OPEN call, but there is no declaration of this method. I think I need the kernel function that syscall 2 open() uses. Can anyone please help me out? Thanks a lot, Jakub _______________________________________________ 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: https://lists.apple.com/mailman/options/darwin-kernel/site_archiver%40lists… This email sent to site_archiver(a)lists.apple.com

1 0

[Q] can I ask your insight for an upcoming project?
by JongAm Park 17 Oct '12

17 Oct '12

Hello, I unsubscribed from these darwin-drivers and darwin-kernel mailing list a few days ago because I didn't work on those stuff anymore, but suddenly I'm in those again. So, I subscribed those two mailing lists again. For my background, I have been a long time, Mac, Windows and Linux programmer in user-space level. I have never touched IOKit or kext drivers before, although I touched OSes from PSOSystem/Windriver to put on a embedded system many many years ago. What I remember now is that most of the user-space libraries and others are useless in the kernel level, and to be able to program in that level, you should know details of H/W, like memory location for specific features etc. So, writing kernel drivers or file systems are like a new animal to me. At work, we are considering a method to bridge a storage device, built totally in-house, connected through PCI-e. Because we want it to be mounted and used just like external HDDs, I thought that we would need to write file system for that. However, yesterday a second thought struck me. Writing file system is complicated although it depends on the features it supports and generality and so on, but instead how about making a kind of driver which bridges the device to HFS+ file system? I believe it will be possible, because we can format USB thumb drives as FAT32, NTFS, HFS+ and so on and mount. I don't think Apple provides HFS+ for USB, HFS+ for HDD, HFS+ for...., for example. So, can you guide me roads in front of us? I noticed some documents like : - Mass Storage Device Driver Programming Guide - I/O Kit fundamentals - Device File Access Guide for Storage Device : but I think it's not for us. This document is about access file storage devices from application program directly. ( but can contain some useful information) - Kernel Frmaework Reference So, do you think it is right path to make such a driver to bridge the device to Apple's existing file system? Are there any documents which explains procedures for simple example projects? Application programs running on Mac are expected to access those files in the in-house device using fope(), fclose(), fread(), fwrite(), etc and other Cocoa framework classes, just like general file access method. Thank you. JongAm Park _______________________________________________ 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: https://lists.apple.com/mailman/options/darwin-kernel/site_archiver%40lists… This email sent to site_archiver(a)lists.apple.com

1 0

Strange MemoryDescriptor behavior
by Maxim Zhuravlev 10 Oct '12

10 Oct '12

Hi, everyone. I'm observing the following strange situation: 1) I need to get a physical address of a part of my kext's code. I do it with the following code: uint64_t ClassName::virtualAddressToPhysicalAddress(const void* address) { IOReturn ioresult = kIOReturnSuccess; uint64_t result; IOMemoryDescriptor* descriptor = IOMemoryDescriptor::withAddressRange((mach_vm_address_t)address, 1, kIODirectionOut, kernel_task); if (descriptor == NULL) return -1; if ((ioresult = descriptor->prepare()) == kIOReturnSuccess) { result = getPhysicalAddressForDescriptor(descriptor); descriptor->complete(); } else { return -1; } descriptor->release(); return result; } uint64_t ClassName::getPhysicalAddressForDescriptor(IOMemoryDescriptor* descriptor) { uint64_t address = -1; IODMACommand* dma = IODMACommand::withSpecification(kIODMACommandOutputHost64, 64, 0, IODMACommand::kMapped, 0, 1); if (!dma) return address; if (kIOReturnSuccess == dma->setMemoryDescriptor(descriptor)) { IODMACommand::Segment64 segments[1]; UInt32 numSegments = 1; UInt64 offset = 0; if (kIOReturnSuccess == dma->gen64IOVMSegments(&offset, segments, &numSegments)) address = segments[0].fIOVMAddr; dma->clearMemoryDescriptor(); } dma->release(); return address; } 2) I need to wire a page in kernel task. I do it with the following code: void* ClassName::allocatePage(uint64_t* physical_address, bool within4GB) { void* result = NULL; IOBufferMemoryDescriptor* descriptor = NULL; if (within4GB) descriptor = IOBufferMemoryDescriptor::inTaskWithPhysicalMask(kernel_task, kIODirectionInOut| kIOMemoryPhysicallyContiguous| kIOMemoryKernelUserShared, page_size, 0x00000000FFFFFFFFull); else descriptor = IOBufferMemoryDescriptor::withOptions( kIODirectionInOut|kIOMemoryPhysicallyContiguous, page_size, page_size); if (descriptor == NULL) return NULL; descriptor->prepare(); *physical_address = getPhysicalAddressForDescriptor(descriptor); result = descriptor->getBytesNoCopy(); if (result) bzero(result, page_size); return result; } 3) I need to call those two subsequently. It used to work and return different physical addresses. But starting OS X 10.8.2 they may return same physical address. I.e. allocatePage may return the same physical address as virtualAddressToPhysicalAddress called for my kext's code address. I can observe it on latest MacBooks (Mid 2012) with OS X 10.8.2. It does work on the same machines with OS X 10.8.1 and earlier. My question is, if I'm doing it wrong from start, or did something change in the way the latest OS X version work with aforementioned hardware? _______________________________________________ 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: https://lists.apple.com/mailman/options/darwin-kernel/site_archiver%40lists… This email sent to site_archiver(a)lists.apple.com

1 0

Re: kext is not authentic; omitting
by Kunal Deo 04 Oct '12

04 Oct '12

_______________________________________________ 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: https://lists.apple.com/mailman/options/darwin-kernel/site_archiver%40lists… This email sent to site_archiver(a)lists.apple.com

1 0

Re: kext is not authentic; omitting
by Shantonu Sen 04 Oct '12

04 Oct '12

_______________________________________________ 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: https://lists.apple.com/mailman/options/darwin-kernel/site_archiver%40lists… This email sent to site_archiver(a)lists.apple.com

1 0

Re: kext is not authentic; omitting
by Kunal Deo 04 Oct '12

04 Oct '12

_______________________________________________ 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: https://lists.apple.com/mailman/options/darwin-kernel/site_archiver%40lists… This email sent to site_archiver(a)lists.apple.com

1 0

Re: kext is not authentic; omitting
by Shantonu Sen 03 Oct '12

03 Oct '12

_______________________________________________ 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: https://lists.apple.com/mailman/options/darwin-kernel/site_archiver%40lists… This email sent to site_archiver(a)lists.apple.com

1 0

kext is not authentic; omitting
by Kunal Deo 03 Oct '12

03 Oct '12

_______________________________________________ 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: https://lists.apple.com/mailman/options/darwin-kernel/site_archiver%40lists… This email sent to site_archiver(a)lists.apple.com

1 0