Three questions about FreeBSD kernel internals
Hello everyone, I've got three small questions about some of the source code in the FreeBSD kernel, based on some of the source code and the manpages. I didn't send it to questions@, because of their technical nature. >>> Question 1: Sleepqueues and turnstiles are allocated on thread creation. Why can't they be allocated when needed? I think the following answer should be quite okay: I suspect because we can't always run sleepq_alloc(). That way we're use we always have a sleepqueue for the process when we need it. We always need these, because otherwise the process is doomed when it can't allocate one. It's also faster I guess, because we don't have to allocate a brand new sleepqueue whenever we walk into a condition on which we have to wait (which happens a real lot, because a real lot of processes sleep). The reason why we don't keep them on the freelist as much as possible, is because we don't have to lock the hash table that much. We've always got one at hand. >>> Question 2: The kernel has three mechanisms to wait for an asynchronous event, namely sema(9), condvar(9) and sleep(9). Strictly, there are only two interfaces, because sema(9) is implemented using mutex(9) and condvar(9). My question is: which interface is the preferred one when writing new code? >>> Question 3: In the file tty_subr.c there is the following comment: | /* | * Allocate an initial base set of cblocks as a 'slush'. | * We allocate non-slush cblocks with each initial tty_open() and | * deallocate them with each tty_close(). | * We should adjust the slush allocation. This can't be done in | * the i/o routines because they are sometimes called from | * interrupt handlers when it may be unsafe to call malloc(). | */ My question is whether the bottom three lines of the comment are still accurate. If I believe the manpage, it's safe to call malloc() in interrupt handlers, if you use M_NOWAIT. I'm not really familiar with older xBSD implementations, but is it true that the kernel couldn't allocate memory in interrupt handlers back then? Yours, -- Ed Schouten <[EMAIL PROTECTED]> WWW: http://g-rave.nl/ pgpsO62PnMEpU.pgp Description: PGP signature
Re: NDIS debugging
I forgot to mention: # uname -a FreeBSD 8.0-CURRENT FreeBSD 8.0-CURRENT #4: Sun Jan 13 12:46:48 BRST 2008[EMAIL PROTECTED]:/usr/src/sys/amd64/compile/GENERICamd64 # uname -m amd64 # I'm using CURRENT as of 11/Jan/08. On Jan 13, 2008 12:52 PM, Diego Giagio <[EMAIL PROTECTED]> wrote: > Hi, > > Recently i've bought a new laptop (Dell Precision M2300) which came > with Intel 4965 Wifi chipset. Since there's no native driver available > yet, I've been trying to get it to work throught Project Evil (NDIS > Windows Driver NETw4x64). > > At first, the driver complained for "no match for > InitializeSListHead". Then I figured it out how to implement that > function (a simple bzero on slist_head, just like ndiswrapper on > Linux) and it stopped complaining. I'm able to kldload NETx4x64_sys.ko > with success, but it stays silent. It just loads two dependent modules > (if_nids.ko and nids.ko) and nothing is printed into dmesg or console. > Also, there's no ndis0 device. > > As far as I can get I could see that DriverEntry()'s module is working > OK as it returns STATUS_SUCCESS on subr_windrv.c's windrv_load(). > > So, the question is: Is there a way to turn on verbose debugging on > the NDIS subsystem? Would be great if I could get some more info on > what's going wrong. > > Thank you. > > -- > DG > ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to "[EMAIL PROTECTED]"
NDIS debugging
Hi, Recently i've bought a new laptop (Dell Precision M2300) which came with Intel 4965 Wifi chipset. Since there's no native driver available yet, I've been trying to get it to work throught Project Evil (NDIS Windows Driver NETw4x64). At first, the driver complained for "no match for InitializeSListHead". Then I figured it out how to implement that function (a simple bzero on slist_head, just like ndiswrapper on Linux) and it stopped complaining. I'm able to kldload NETx4x64_sys.ko with success, but it stays silent. It just loads two dependent modules (if_nids.ko and nids.ko) and nothing is printed into dmesg or console. Also, there's no ndis0 device. As far as I can get I could see that DriverEntry()'s module is working OK as it returns STATUS_SUCCESS on subr_windrv.c's windrv_load(). So, the question is: Is there a way to turn on verbose debugging on the NDIS subsystem? Would be great if I could get some more info on what's going wrong. Thank you. -- DG ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to "[EMAIL PROTECTED]"
mutex lock for filesystem list.
Hi all, I'm trying to write a patch to add a file to the linprocfs. This file provides the list of filesystems present in the kernel. Though I was able to create the file and traverse the list, it's clearly unsafe. I would like to know wich semaphore I should lock cause I can't find it in the kernel code. In addition I'd like to know if there is a book similar to the Understanding the Linux kernel, but for freebsd. What is the best information source to get introduced in the freebsd kernel programming (apart of the code itself)? Thanks in advance ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: mutex lock for filesystem list.
* Fernando Apesteguía <[EMAIL PROTECTED]> wrote: > I'm trying to write a patch to add a file to the linprocfs. This file > provides the list of filesystems present in the kernel. > > Though I was able to create the file and traverse the list, it's > clearly unsafe. I would like to know wich semaphore I should lock > cause I can't find it in the kernel code. > > In addition I'd like to know if there is a book similar to the > Understanding the Linux kernel, but for freebsd. What is the best > information source to get introduced in the freebsd kernel programming > (apart of the code itself)? You're probably looking through the mountlist? Then you need to make sure you lock the mountlist_mtx. See src/sys/kern/vfs_mount.c. -- Ed Schouten <[EMAIL PROTECTED]> WWW: http://g-rave.nl/ pgpGDJHYOuLMX.pgp Description: PGP signature
Re: mutex lock for filesystem list.
And of course I forgot to answer all the questions. * Fernando Apesteguía <[EMAIL PROTECTED]> wrote: > In addition I'd like to know if there is a book similar to the > Understanding the Linux kernel, but for freebsd. What is the best > information source to get introduced in the freebsd kernel programming > (apart of the code itself)? I guess most people would say "The Design and Implementation of the FreeBSD Operating System", by McKusick and Neville-Neil: http://www.amazon.com/Design-Implementation-FreeBSD-Operating-System/dp/0201702452 Yours, -- Ed Schouten <[EMAIL PROTECTED]> WWW: http://g-rave.nl/ pgpumS3I1zaDJ.pgp Description: PGP signature
Re: mutex lock for filesystem list.
Ed Schouten wrote: And of course I forgot to answer all the questions. * Fernando Apesteguía <[EMAIL PROTECTED]> wrote: In addition I'd like to know if there is a book similar to the Understanding the Linux kernel, but for freebsd. What is the best information source to get introduced in the freebsd kernel programming (apart of the code itself)? I guess most people would say "The Design and Implementation of the FreeBSD Operating System", by McKusick and Neville-Neil: http://www.amazon.com/Design-Implementation-FreeBSD-Operating-System/dp/0201702452 but be warned that th at book was written part way through the SMP rewrite so a lot has changed... Dr McKusick said the words "Next edition" the other day but I think it's still just a glimmer in his eye. Yours, ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Nvidia Driver w/RELENG_7
Hi, Does anyone have the nvidia video driver working with releng 7? What I am seeing is that the kernel module compiles and loads fine, but it will not detect any video card. I've tried removing the agp device from my BSD kernel and using nvidia's. -- Ali Mashtizadeh علی مشتی زاده ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: Nvidia Driver w/RELENG_7
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Ali Mashtizadeh wrote: > Hi, > > Does anyone have the nvidia video driver working with releng 7? What I am > seeing is that the kernel module compiles and loads fine, but it will not > detect any video card. I've tried removing the agp device from my BSD kernel > and using nvidia's. I'm using xf86-video-nv, I think you are referring the binary driver, which is not this one? - -- Xin LI <[EMAIL PROTECTED]> http://www.delphij.net/ FreeBSD - The Power to Serve! -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.4 (FreeBSD) iD8DBQFHisBAi+vbBBjt66ARAob/AKCQxMuw0DvAZixDm+8qvU+lGpLNhwCfdelk at4T0KRlKFKfMq2sQiCElcE= =oL2t -END PGP SIGNATURE- ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: Nvidia Driver w/RELENG_7
On Mon, 14 Jan 2008, Ali Mashtizadeh wrote: > Does anyone have the nvidia video driver working with releng 7? What > I am seeing is that the kernel module compiles and loads fine, but it > will not detect any video card. I've tried removing the agp device > from my BSD kernel and using nvidia's. Load the module in the loader rather than after the kernel has booted. acpi_video can also grab the device I believe.. I am running current (ie 8) and I ran -current when it was 7.x and loading it from the loader has always worked (modulo incompatible source changes :) -- Daniel O'Connor software and network engineer for Genesis Software - http://www.gsoft.com.au "The nice thing about standards is that there are so many of them to choose from." -- Andrew Tanenbaum GPG Fingerprint - 5596 B766 97C0 0E94 4347 295E E593 DC20 7B3F CE8C signature.asc Description: This is a digitally signed message part.
Re: Nvidia Driver w/RELENG_7
On Sun, Jan 13, 2008 at 05:22:46PM -0800, Ali Mashtizadeh wrote: > Hi, > > Does anyone have the nvidia video driver working with releng 7? What I am > seeing is that the kernel module compiles and loads fine, but it will not > detect any video card. I've tried removing the agp device from my BSD kernel > and using nvidia's. > > -- > Ali Mashtizadeh > علی مشتی زاده If you are sure that installed version of nvidia-driver supports your card, try putting nvidia_load="YES" in /boot/loader.conf. I guess your problem is vgapci attaching to device, leaving nvidia module without hardware to attach, like below: [EMAIL PROTECTED]:5:0:0: class=0x03 card=0x40161682 chip=0x040210de rev=0xa1 hdr=0x00 vendor = 'Nvidia Corp' device = 'GeForce 8600 GT' class = display subclass = VGA HTH, Yuri ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: Nvidia Driver w/RELENG_7
it work right for me: [EMAIL PROTECTED]:1:0:0: #uname -srm FreeBSD 7.0-RC1 i386 ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: netgraph question
Subhash Gopinath wrote: Thanks, looks interesting. But I was looking at processing the packets in userspace. Sorry I didn't mention it clearly. Ah ok. I didn't get that from your initial email. Have you looked at the firewall (ipfw and/or pf) code at all? I believe you can use mechanisms like divert sockets (man 4 divert) to pass packets up from the kernel to userspace for processing, and then reinject the packets into the stack if they pass whatever criteria is required. I'm sure there are other mechanisms for getting packets up into userspace as well, but firewall code is probably a good place to start looking for ideas. Thanks, -Subhash On Jan 11, 2008 10:32 PM, Lawrence Stewart <[EMAIL PROTECTED]> wrote: Hi Subhash, Subhash Gopinath wrote: Hello folks, I am looking at writing an application program to tap certain ipv6 packets (say icmpv6) using netgraph. The application has to do some processing, before kernel can proceed with those packets. I have vaguely understood netgraph, and I see that I need a ng_socket node in the application, an ng_bpf node, and an ng_ether or ng_iface node in the kernel. My question is. would I need to create such nodes for each interface. Then it becomes unscalable.. Can I have just one socket, bpf, iface node that can tap icmpv6 packets on all interfaces? The PFIL(9) interface might also be of interest to you. If all you need to do is packet interception and then allow/deny packets based on the results of some processing, PFIL might be the way to go. We wrote some code (SIFTR [1]) which uses PFIL in a similar capacity and you may want to refer to it as an example. Cheers, Lawrence [1] http://caia.swin.edu.au/urp/newtcp/tools.html Cheers, Lawrence ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to "[EMAIL PROTECTED]"
