October 2014 - kernelnewbies@kernelnewbies.org

doubt about interrupts and spinlocks
by Oscar Salvador Oct. 27, 2014

Oct. 27, 2014

Hi! I have a doubt about interrupts and spin locks. I'm not sure if I'm in the good way: When an interrupt is raised, the kernel (through do_IRQ and etc) acknowledges the interrupt and locks this IRQ taking a spin lock, so any other cpu can not handle this interrupt, and then the kernel executes the handler, if there is a handler available for that interrupt. When we are on the handler interrupt , maybe we need to lock the cpu till we finish( if we are in a network driver, may we need to process some TX/RX packets ), so if we have to lock the code we can do a spin_lock(). But with a spin_lock, even other CPU's can execute part of your code(code that is not part of the handler interrupt). For example: (this is an example in a network device driver) The system has to transmit a packet, so network subsystem calls ndo_start_xmit. Inside this function we take a lock disabling the interrupts(spin_lock_irqsave) because we have to do some private things, but suddenly an interrupt is raised and we go to the interrupt handler, we take a spin_lock, we do some work (processing some RX/TX packets), and exites from the interrupt handler, and then we come back to the ndo_start_xmit when we were. As far I could understand( I have read http://www.makelinux.net/ldd3/chp-5-sect-5 ), spin_lock_irqsave only disables the interrupts for the local cpu(where the code is executed), so I think that another cpu took the handler interrupt. The problem is if I'm working with irq disabled, because I have to handle a private structure and suddenly interrupts is taking place, the interrupt can manipulate the code that I'm trying to keep safe. Is there any advice? Maybe that is happened because the code within my critical section on ndo_start_xmit has to be atomic? Otherwise it can go to sleep and I can lose my lock? Thank you very much Oscar Salvador

3 3

Subsystem communication
by Stefan Wahren Oct. 27, 2014

Oct. 27, 2014

Hi, currently i want to implement a driver for the power subsystem of the Freescale i.MX28 (ARM9). On the one hand the driver should trigger the registering of the underlying components like regulators and on the other hand it should take care of the incoming interrupts and provide the results to the underlying components. I think the first task can easily done by of_platform_populate(). But for the second task i'm not sure how to achieve a clean, mostly readonly access from the regulator drivers to the power subsystem driver. I have read about exporting symbols and RCU, but i'm not sure what's preferred in that situation. Best regards Stefan

1 1

I remove the audit and selinux from kernel and start the new kernel in Centos, but i can't bring up the network device, Why?
by sizel Oct. 27, 2014

Oct. 27, 2014

4 3

Kconfig options and make command line
by Peter Senna Tschudin Oct. 26, 2014

Oct. 26, 2014

I was trying to compile the file: drivers/staging/rtl8723au/core/rtw_ap.c but $ make drivers/staging/rtl8723au/core/rtw_ap.o doesn't work as there is no Makefile at core directory. Doing: $ make drivers/staging/rtl8723au/ doesn't work also as rtw_ap.c is only compiled under certain Kconfig conditions due this line at the Makefile: r8723au-$(CONFIG_8723AU_AP_MODE) += core/rtw_ap.o So is there a way to compile rtw_ap.o by just passing arguments to make? -- Peter

3 3

Re: i2c-dev
by Paul Bolle Oct. 26, 2014

Oct. 26, 2014

On Wed, 2014-10-22 at 13:27 -0400, John de la Garza wrote: > On Wed, Oct 22, 2014 at 06:35:02PM +0200, Paul Bolle wrote: > > Having looked into this a bit, but failing to come up with something > > helpful, I decided to ask some additional info. > > > > On Tue, 2014-10-21 at 21:54 -0400, John de la Garza wrote: > > > I have my kernel > > > > Which version (roughly)? > > Linux vega 3.18.0-rc1+ #71 SMP Wed Oct 22 00:49:21 EDT 2014 x86_64 GNU/Linux > > > > > built with dev-i2c and i2c_i801 > > > > .config, please? > > > see below > > > > but do not have > > > a /dev entry. > > > > > > lspci shows this: > > > 00:1f.3 SMBus: Intel Corporation 5 Series/3400 Series Chipset SMBus Controller (rev 06) > > > > What are the PCI vendor and device codes? See "lspci -nn". > > > 00:1f.3 SMBus [0c05]: Intel Corporation 5 Series/3400 Series Chipset SMBus Controller [8086:3b30] (rev 06) > > > dmesg shows this: > > > [ 1.566003] i801_smbus 0000:00:1f.3: enabling device (0000 -> 0003) I'm afraid I can't really help you here, sorry. That is, except for the obvious question whether this is a regression (ie, do releases v3.17 or earlier show something in /sys/bus/i2c/devices/?) or a configuration issue (perhaps a distribution provided kernel does show what you want). Hope this helps, Paul Bolle

2 1

sendmmsg() Expected Behavior?
by Hei Chan Oct. 24, 2014

Oct. 24, 2014

Hi, I am on CentOS 6.5 with kernel version kernel-2.6.32-431.el6. I was trying to find the implementation of recvmmsg(), and hopefully, I am not looking at the wrong source ~/rpmbuild/BUILD/kernel-2.6.32-431.el6/linux-2.6.32-431.el6.x86_64/net/socket.c. If I am, please kindly point me to the right source. int __sys_recvmmsg(int fd, struct mmsghdr __user *mmsg, unsigned int vlen, unsigned int flags, struct timespec *timeout) { int fput_needed, err, datagrams; struct socket *sock; struct mmsghdr __user *entry; struct compat_mmsghdr __user *compat_entry; struct msghdr msg_sys; struct timespec end_time; if (timeout && poll_select_set_timeout(&end_time, timeout->tv_sec, timeout->tv_nsec)) return -EINVAL; datagrams = 0; sock = sockfd_lookup_light(fd, &err, &fput_needed); if (!sock) return err; err = sock_error(sock->sk); if (err) goto out_put; entry = mmsg; compat_entry = (struct compat_mmsghdr __user *)mmsg; while (datagrams < vlen) { /* * No need to ask LSM for more than the first datagram. */ if (MSG_CMSG_COMPAT & flags) { err = __sys_recvmsg(sock, (struct msghdr __user *)compat_entry, &msg_sys, flags, datagrams); if (err < 0) break; err = __put_user(err, &compat_entry->msg_len); ++compat_entry; } else { err = __sys_recvmsg(sock, (struct msghdr __user *)entry, &msg_sys, flags, datagrams); if (err < 0) break; err = put_user(err, &entry->msg_len); ++entry; } if (err) break; ++datagrams; if (timeout) { ktime_get_ts(timeout); *timeout = timespec_sub(end_time, *timeout); if (timeout->tv_sec < 0) { timeout->tv_sec = timeout->tv_nsec = 0; break; } /* Timeout, return less than vlen datagrams */ if (timeout->tv_nsec == 0 && timeout->tv_sec == 0) break; } /* Out of band data, return right away */ if (msg_sys.msg_flags & MSG_OOB) break; } out_put: fput_light(sock->file, fput_needed); if (err == 0) return datagrams; if (datagrams != 0) { /* * We may return less entries than requested (vlen) if the * sock is non block and there aren't enough datagrams... */ if (err != -EAGAIN) { /* * ... or if recvmsg returns an error after we * received some datagrams, where we record the * error to return on the next call or if the * app asks about it using getsockopt(SO_ERROR). */ sock->sk->sk_err = -err; } return datagrams; } return err; } Let's say there are 2 packets at the socket, and I try: timespec t = {0, 0}; recvmmsg(fd, mmsg, vlen, 0, &t); Then, because of these lines: if (timeout->tv_nsec == 0 && timeout->tv_sec == 0) break; recvmmsg() will just return 1 packet instead of 2. Am I right? It seems defecting the one of the main purposes of recvmmsg() -- trying to reduce # of system calls when there are packet queued. Maybe I miss something? I created a small test case that it waits for epoll callback, sleeps for 2-5 seconds and then calls recvmmsg() with timeout=0 (not null; just like above). Then, I found that recvmmsg() doesn't return all the packets that queued in 1 single call.

2 3

Documentation/DMA-API-HOWTO.txt 2
by Peter Senna Tschudin Oct. 24, 2014

Oct. 24, 2014

This was edited in 2010, but the original is more than 9 years old, and still there are some drivers using virt_to_bus() and bus_to_virt(). -- // -- All drivers should be using these interfaces with no exceptions. It is planned to completely remove virt_to_bus() and bus_to_virt() as they are entirely deprecated. Some ports already do not provide these as it is impossible to correctly support them. -- // -- Files calling virt_to_bus: ./drivers/atm/ambassador.c ./drivers/atm/firestream.c ./drivers/atm/horizon.c ./drivers/atm/zatm.c ./drivers/block/swim3.c ./drivers/isdn/hisax/netjet.c ./drivers/media/pci/zoran/zoran_device.c ./drivers/media/pci/zoran/zoran_driver.c ./drivers/net/appletalk/ltpc.c ./drivers/net/ethernet/amd/au1000_eth.c ./drivers/net/ethernet/apple/bmac.c ./drivers/net/ethernet/apple/mace.c ./drivers/net/ethernet/dec/tulip/de4x5.c ./drivers/net/ethernet/i825xx/82596.c ./drivers/net/ethernet/sgi/ioc3-eth.c ./drivers/net/irda/au1k_ir.c ./drivers/net/irda/donauboe.c ./drivers/net/wan/cosa.c ./drivers/net/wan/lmc/lmc_main.c ./drivers/net/wan/z85230.c ./drivers/scsi/a2091.c ./drivers/scsi/a3000.c ./drivers/scsi/advansys.c ./drivers/scsi/gvp11.c ./drivers/scsi/mvme147.c ./drivers/staging/comedi/drivers/das1800.c ./drivers/staging/comedi/drivers/dt282x.c ./drivers/staging/comedi/drivers/ni_at_a2150.c ./drivers/staging/comedi/drivers/ni_labpc_isadma.c ./drivers/staging/comedi/drivers/pcl812.c ./drivers/staging/comedi/drivers/pcl816.c ./drivers/staging/comedi/drivers/pcl818.c ./drivers/staging/netlogic/xlr_net.c ./drivers/staging/slicoss/slicoss.c ./drivers/tty/serial/cpm_uart/cpm_uart_cpm2.c ./drivers/vme/bridges/vme_ca91cx42.c Files calling bus_to_virt: ./drivers/atm/ambassador.c ./drivers/atm/firestream.c ./drivers/atm/horizon.c ./drivers/atm/zatm.c ./drivers/isdn/hisax/netjet.c ./drivers/media/pci/zoran/zoran_driver.c ./drivers/net/ethernet/apple/bmac.c ./drivers/net/ethernet/apple/mace.c ./drivers/scsi/BusLogic.c ./drivers/scsi/dpt_i2o.c ./drivers/staging/netlogic/xlr_net.c Should the 'with no exceptions' be enforced? -- Peter

2 1

Documentation/DMA-API-HOWTO.txt
by Peter Senna Tschudin Oct. 24, 2014

Oct. 24, 2014

There is a portion of DMA-API-HOWTO.txt that was not updated for a long time. -- // -- This means specifically that you may _not_ use the memory/addresses returned from vmalloc() for DMA. It is possible to DMA to the _underlying_ memory mapped into a vmalloc() area, but this requires walking page tables to get the physical addresses, and then translating each of those pages back to a kernel address using something like __va(). [ EDIT: Update this when we integrate Gerd Knorr's generic code which does this. ] -- // -- Was Gerd Knorr's generic code integrated? -- Peter

1 0

/drivers/staging/TODO doc and kernel tasks question
by Jaime Arrocha Oct. 23, 2014

Oct. 23, 2014

Good day, The TODO document has the addresses of where to send a patch, but if you run get_maintainer.pl it gives me a different person to address it, gregkh for example. I have sent a patch to the persons on the TODO list, should I still copy to the maintainer in charge too? CC maybe? Another question, apart from the /drivers/staging and bugtracker, is there some where else to look for bug info? The mailing list? I just want to get the complete picture. Last question, I understand that if there is a bug, most of the time they tell you to try to reproduce it on the latest kernel. If it doesn't occur then everything is good and I should look for another one. But who is in charge of porting a bug fix like these to old kernels? How can I help? Or I, as a newbie, should focus more on the /drivers/staging and bugtracker? I want to thank in advance for your time and please let me know if I can read about all this in some website.

2 2

Non-maskable interrupt
by J.Hwan Kim Oct. 23, 2014

Oct. 23, 2014

Hi, everyone In my system, NMI intrrupt counter is increased continuously, which I found in /proc/interrupts. How can I find the reasons of NMI interrupts? Best Regards Kim.

4 5