February 2008 - Darwin-kernel

Re: see if a mutex lock is taken?
by Greg 12 Feb '08

12 Feb '08

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com On Feb 11, 2008, at 3:42 PM, mm w wrote: hi Greg, yep these mechanisms exist, this is the first things that you implement when you write a lock system no matter on what: islockable and islocked try to understand, what you do (if you use a lck object (locked)) if your lock object has been created, a lock has been aquired when you create a lock object, the lock system test it for you Thanks! - Greg 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.… -mmw Hi mmw, thanks for your reply. I was specifically wondering how I could do this from a kernel extension? I don't see a header in Kernel.framework that contains any try_lock functions for lck_mtx_t. Hmm... I've found this in the sources to xnu, but I can't find this in the Kernel.framework. Perhaps I should have mentioned that I'm writing a kernel extension. Does this exist for kernel extensions, and if so how do I access it? Right now I'm just linking against Kernel.framework, which doesn't seem to have this header. This email sent to site_archiver(a)lists.apple.com

1 0

Re: see if a mutex lock is taken?
by mm w 11 Feb '08

11 Feb '08

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com Dkim-signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=gamma; h=domainkey-signature:received:received:message-id:date:from:to:subject:cc:in-reply-to:mime-version:content-type:content-transfer-encoding:content-disposition:references; bh=ueVL5EsBhgMTf+MchC/M2+vvVGnJZrQVJeuePKIx52Q=; b=L/j9t75Wv/ct+gl5N/XZt0Vjq1fmWreslesiiAqPuK44Eo4icw+CV1CZOxd+Y2pU2hIRQ0dqZZSWQ9nCZyCsR9lYrvnH/olQ9P6zf7Hz5A4OQSegYEzetIQVhE1FHygT0/UE+juMGxc/jmKgGcNWETtbMSayBOsVNF6sG/ufFIU= Domainkey-signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=message-id:date:from:to:subject:cc:in-reply-to:mime-version:content-type:content-transfer-encoding:content-disposition:references; b=f7URXlwCMjSXCRgXu1pz8aHMwDdxIrDvflDbz/hcTLVaMOQKRnmNPXP0t21RA9GxnuSs1osEtQYb62/5TWtUDjoLI0TcwkyIZQ2ZSIjA9U/wytJxKoOy7rDzzXaEAjZJck8gyRcwFqp752B7k4pyPCrKTsBSSck8TOzphZmvbvs= hi Greg, yep these mechanisms exist, this is the first things that you implement when you write a lock system no matter on what: islockable and islocked try to understand, what you do (if you use a lck object (locked)) if your lock object has been created, a lock has been aquired when you create a lock object, the lock system test it for you -mmw On Feb 11, 2008 12:01 PM, Greg <greg(a)kinostudios.com> wrote: > > On Feb 11, 2008, at 2:49 PM, mm w wrote: > > > hi greg, > > > > lock.h > > > > extern boolean_t mutex_try( > > mutex_t *mutex); > > > > whatever it depends of your xnu version > > > > -mmw > > Hmm... I've found this in the sources to xnu, but I can't find this in > the Kernel.framework. Perhaps I should have mentioned that I'm > writing a kernel extension. Does this exist for kernel extensions, > and if so how do I access it? Right now I'm just linking against > Kernel.framework, which doesn't seem to have this header. > > Thanks! > > - Greg > -- -mmw _______________________________________________ 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: see if a mutex lock is taken?
by Greg 11 Feb '08

11 Feb '08

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com On Feb 11, 2008, at 2:49 PM, mm w wrote: hi greg, lock.h extern boolean_t mutex_try( mutex_t *mutex); whatever it depends of your xnu version 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.… -mmw Hmm... I've found this in the sources to xnu, but I can't find this in the Kernel.framework. Perhaps I should have mentioned that I'm writing a kernel extension. Does this exist for kernel extensions, and if so how do I access it? Right now I'm just linking against Kernel.framework, which doesn't seem to have this header. This email sent to site_archiver(a)lists.apple.com

1 0

Re: see if a mutex lock is taken?
by Greg 11 Feb '08

11 Feb '08

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com On Feb 10, 2008, at 10:26 PM, Michael Smith wrote: 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.… The answer is meaningless unless you take the lock in the case where it's not already owned; this is what the 'try' functions do. If you really are just 'sniffing' the lock, call the 'try' function, and if it returns success, release the lock. But I'd think long and hard about what you think you can infer about anything just because the lock wasn't held when you tried it. (If you're just introspecting, then ignore the above caution.) Which 'try' function? The only one I was able to find was in lock_set.h (lock_try) which takes a lock_set_t, not a lck_mtx_t, and I don't know if those can somehow be used together. The only reason that I'm asking is because I was worried about having parts of the kernel wait too long if another part is taking too long to accomplish a task. Right now I think that it's alright if it does wait, so I just have it take the lock, but I was curious if this functionality existed in the kernel. This email sent to site_archiver(a)lists.apple.com

1 0

Re: see if a mutex lock is taken?
by Michael Smith 11 Feb '08

11 Feb '08

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com On Feb 10, 2008, at 1:18 PM, Greg wrote: Greg, = Mike _______________________________________________ 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.… Is there any way to check and see if a lck_mtx_t has been taken? (without modifying the lock itself). The answer is meaningless unless you take the lock in the case where it's not already owned; this is what the 'try' functions do. If you really are just 'sniffing' the lock, call the 'try' function, and if it returns success, release the lock. But I'd think long and hard about what you think you can infer about anything just because the lock wasn't held when you tried it. (If you're just introspecting, then ignore the above caution.) This email sent to site_archiver(a)lists.apple.com

1 0

Re: gssd-agent: Minor error <1> Unknown Error Code: 19777
by Rick Macklem 10 Feb '08

10 Feb '08

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com I realized I didn't say anything useful the last time, so I thought I'd post again with what I observed related to this error. All this assumes that the user with uid==502 is logged in and has a valid TGT in the credentials cache: - upcalls with uid==0 would fail. This makes sense, but since I needed system Ops related to NFSv4 such as Renew to work, I mapped uid 0 to the uid of the user that did the mount (ie 502 in this case). --> when uid 0 was replaced with 502 in the upcall, it worked fine. - I still saw some of these errors for uid == -2. (I presume that some daemon is poking at the file system, but haven't yet tracked that out. Someone was nice enough to email me a dtrace script to do this, but I haven't gotten around to using it yet.) To see if it was caused by the wrong uid, I tried mapping the uid of -2 to 502, like I had done for the system Ops. (Wouldn't make sense to do this for production code, it was just an experiment.) --> Still failed with the error, even though the upcall had the "correct" uid as argument. So, it seems that some processes/threads can't authenticate even if the uid is correct. I suspect it has something to do with accessing the credentials cache (which I understand lives in VM on Leopard). When it comes to control of access to a VM object on the Mach side of things, I know diddly about it, so this is just a wild guess. Hopefully this might be of some use? rick _______________________________________________ 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

see if a mutex lock is taken?
by Greg 10 Feb '08

10 Feb '08

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.… Is there any way to check and see if a lck_mtx_t has been taken? (without modifying the lock itself). This email sent to site_archiver(a)lists.apple.com

1 0

Re: Stop the Ethernet interface from going down
by mm w 07 Feb '08

07 Feb '08

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com Dkim-signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=gamma; h=domainkey-signature:received:received:message-id:date:from:to:subject:in-reply-to:mime-version:content-type:content-transfer-encoding:content-disposition:references; bh=2JvzIGeQAyLY+Dj4TG/GRZQLtRzybTjsrWXDGm37f40=; b=B31fzHpByYJJNxR4i3i26/MR6tq68eIUtL6/zjgNEMGpNgIeV7Gg5GZvcaKWEFtvY71VFjrv0OTWnYtvnUVg08aIDzKVhqjUi8iR+l7z8rdqIWmv4KY9A86g91mdOOP+XQf6Q3xxm6+UEEFQmGYxz6vTIwuZBC5q63+bnpEndvQ= Domainkey-signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=message-id:date:from:to:subject:in-reply-to:mime-version:content-type:content-transfer-encoding:content-disposition:references; b=BKZ+OFxdOkv3ml8UxckFIvxcE/YD6IExAOhfTVEVrpbAKPA75HipaBkzeFm+hApnsk00TLN+2rzkikzcBDLb4A6fXPLRW2BYyaPqMlZczSoE9VJ7qQmmKPSQXR4ttUiQep4+3AdlUOVGyfwdEgGhEMgSGBSl5dGYn7x3QQG5vdg= you should read: "idk if you found a solution, "you cannot control it " -mmw On Feb 7, 2008 5:58 PM, mm w <openspecies(a)gmail.com> wrote: > Hi Rakesh, I dont know if you find a solution I pretty sure yes, > as i said Shantonu, you can control it or try to do something like > this is called in French > "usine a gaz", you cannot face to every cases, if a task/request > failed, the informations are not losed, you have to create a > data/request queue: > > "if last task not satisfy keep it and retry" > > imagine a common company internet-network, with a ton of packet > collision and data drop > > imagine now that your network interface is a mobile phone, > it s quite the same thing, > > when you speak to someone and in the middle of your conversation you > lose the connection, you can recall and continue your conversation > > -mmw > > > On Jan 29, 2008 4:02 PM, Shantonu Sen <ssen(a)apple.com> wrote: > > What happens if the user unplugs their wired ethernet, or they drop out of > > range of their wireless network? The link layer can have any number of > > unforeseen issues, independent of sleep/wake. You NKE should be able to deal > > with errors of this nature, which should also address sleep/wake. > > > > > > > > > > Shantonu > > > > Sent from my MacBook > > > > > > On Jan 29, 2008, at 6:56 AM, Rakesh Singhal wrote: > > I am working on NKE(Network kernel extension). During sleep, the ethernet > > interface goes down and due to it, I am facing some problem. I have to > > complete my current request of data transfer and then it can go down. Is it > > possible to hold it for sometime(some miliseconds) and complete my > > requirement and then let it go down? > > > > Regards > > > > On Jan 29, 2008 7:53 PM, Shantonu Sen <ssen(a)apple.com> wrote: > > > > > > > > > To what end? What are you trying to do? > > > > > > > > > > > > > > > Shantonu > > > > > > > > > Sent from my MacBook > > > > > > > > > > > > > > > > > > On Jan 29, 2008, at 1:47 AM, Rakesh Singhal wrote: > > > > > > > > > > > > > > > Hi all > > > > > > During sleep, the Ethernet interface goes down. Is it possible to hold the > > Ethernet interface for sometime? > > > > > > Thanks in advance > > > > > > With warm regards > > > Rakesh Singhal _______________________________________________ > > > 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/ssen%40apple.com > > > > > > This email sent to ssen(a)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/openspecies%40gmail.com > > > > This email sent to openspecies(a)gmail.com > > > > > > -- > -mmw > -- -mmw _______________________________________________ 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: Stop the Ethernet interface from going down
by mm w 07 Feb '08

07 Feb '08

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com Dkim-signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=gamma; h=domainkey-signature:received:received:message-id:date:from:to:subject:in-reply-to:mime-version:content-type:content-transfer-encoding:content-disposition:references; bh=MaLa3tMRexLDA/FgHrjLyy7w1wUUkIMcc7oCle3tIzA=; b=rqAfV+2jELq9KHM/UJ+Bv02Cm+x110+4GrqO70J35X0ufH+6IcL1gZlbPWMDtpDfdF0cmrTuDpeAvDk3Zcl/s3DTwBt+aGG7iP+wSKS9ukrYbbkoL6VUSDmbg5QQsrrDZNGJ845LRCRyd08ggjKitPiyni/c/GKQFC/QYPVzYqU= Domainkey-signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=message-id:date:from:to:subject:in-reply-to:mime-version:content-type:content-transfer-encoding:content-disposition:references; b=k1f7yvPEycY5+uoKL9j98FVHDl5q9ABwsu7uHb6bZ33zfK55drfgqBJO8Rga/zz3aYeM2VjnP6462qJYP0XzOPM7Ka37jN4yOUo8gtPLzMqkK/oZBr41f/S58Au2hqkuAgp2/HzDB2oz9KGTfUPRTpyGCCSxubf226AfTLXJUVY= Hi Rakesh, I dont know if you find a solution I pretty sure yes, as i said Shantonu, you can control it or try to do something like this is called in French "usine a gaz", you cannot face to every cases, if a task/request failed, the informations are not losed, you have to create a data/request queue: "if last task not satisfy keep it and retry" imagine a common company internet-network, with a ton of packet collision and data drop imagine now that your network interface is a mobile phone, it s quite the same thing, when you speak to someone and in the middle of your conversation you lose the connection, you can recall and continue your conversation -mmw On Jan 29, 2008 4:02 PM, Shantonu Sen <ssen(a)apple.com> wrote: > What happens if the user unplugs their wired ethernet, or they drop out of > range of their wireless network? The link layer can have any number of > unforeseen issues, independent of sleep/wake. You NKE should be able to deal > with errors of this nature, which should also address sleep/wake. > > > > > Shantonu > > Sent from my MacBook > > > On Jan 29, 2008, at 6:56 AM, Rakesh Singhal wrote: > I am working on NKE(Network kernel extension). During sleep, the ethernet > interface goes down and due to it, I am facing some problem. I have to > complete my current request of data transfer and then it can go down. Is it > possible to hold it for sometime(some miliseconds) and complete my > requirement and then let it go down? > > Regards > > On Jan 29, 2008 7:53 PM, Shantonu Sen <ssen(a)apple.com> wrote: > > > > > > To what end? What are you trying to do? > > > > > > > > > > Shantonu > > > > > > Sent from my MacBook > > > > > > > > > > > > On Jan 29, 2008, at 1:47 AM, Rakesh Singhal wrote: > > > > > > > > > > Hi all > > > > During sleep, the Ethernet interface goes down. Is it possible to hold the > Ethernet interface for sometime? > > > > Thanks in advance > > > > With warm regards > > Rakesh Singhal _______________________________________________ > > 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/ssen%40apple.com > > > > This email sent to ssen(a)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/openspecies%40gmail.com > > This email sent to openspecies(a)gmail.com > -- -mmw _______________________________________________ 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: gssd-agent: Minor error <1> Unknown Error Code: 19777
by Rick Macklem 07 Feb '08

07 Feb '08

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com On Wed, 6 Feb 2008, Terry Lambert wrote: Hi list, I'm running a cluster of Mac's with kerberized nfs, and I keep getting this rather cryptic error message in my log: We are running SGE (sun grid engine), and the above error prevents us from writing to nfs directories. So far, our solution is to reboot the machine. I'm wondering if anyone has had similar experiences, and if anyone knows what error code 19777 actually means? Here's an answer from the maintainer: You get the message if you try to access a kerberized NFS mount without a kerberos ticket - in this case it was user 501 - which looks like a local UID - not a network one. So, just seeing a valid TGT isn't enough. Don't know if this helps or just adds to the confusion? rick ps: Of course, if you don't have a valid TGT for uid==501, then nothing is going to work. Leopard expects user credentials and doesn't use host based principals like (root/client.domain@REALM) in a keytab file, as far as I understand it. (Solaris will use root/client.dns.domain@REALM for root accesses, if that exists in the client machine's keytab file.) _______________________________________________ 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 Feb 6, 2008, at 2:13 AM, Roger Herikstad wrote: Feb 6 17:31:52 work03 gssd-agent[227]: Error returned by svc_mach_gss_init_sec_context: Feb 6 17:31:52 work03 gssd-agent[227]: Major error <1> Unspecified GSS failure. Minor code may provide more information Feb 6 17:31:52 work03 gssd-agent[227]: Minor error <1> Unknown Error Code: 19777 Feb 6 17:31:52 work03 gssd-agent[227]: nfs client Kerberos: head.neuralc:/Volumes/Xraid/XUsers, uid=501 - Unknown Error Code: 19777 There's a known bug that prevents the message from being printed, but error code 19777 is a kerberos library error that actually corresponds to the text: "Can't display user interface from this environment". The message implies that the kerberos library attempted to pop up a window to obtain a kerberos login because you're lacking a ticket. You'll normally get a kerberos ticket when you log in through the loginwindow or the screen saver. In good old nfs, a local uid is all there is. (From playing with the gssd doing upcalls from my client, the uid argument in the upcall doesn't seem to much matter, anyhow.) I think the error is generated when the process trying the NFS operation is somehow not associated with the gssd/credential cache that has the correct TGT. (I believe Quinn's answer sounds like a good explanation of that.) I know that I can get that error to occur by doing an upcall for uid==502 when the user with uid==502 has a valid TGT (whereas other upcalls for uid==502 work fine). This email sent to site_archiver(a)lists.apple.com

1 0