- kernelnewbies@kernelnewbies.org - Kernelnewbies Mailing Lists

On common struct clock in clock framework.
by mind entropy June 30, 2013

June 30, 2013

Hi, I am reading through mini2440 code and understanding the clocking code. I am getting confused as to whether it uses the common clocking framework. In https://github.com/torvalds/linux/blob/master/arch/arm/plat-samsung/include… they define a struct clk. Even in OMAP (http://processors.wiki.ti.com/index.php/Clock_Framework_User_Guide#Software…) they have a separate struct clk definition. In the common clocking framework shouldn't they be using the common struct clk rather than have their own versions of struct clk? Thanks in advance.

1 0

Kernel Linux updating
by Mylene Josserand June 29, 2013

June 29, 2013

Hi everyone, I would like to know how the Linux Kernel are updated. I know that there is the long term kernels, the last stables and the mainline. First, what is the real difference between stable and longterm ? I see in the Linux Kernel website that the date of the long term (and the revision number) is changing so I was thinking that there are some updates on it, right ? For example, when an important bug has been fixed, is it fixed in the new release kernel only ? Or is it applied on old kernels ? Only the long term ? All ? So if I am using the long term 3.4.49 for example (and the current is 3.4.51), I can just update the 3.4.49 to get the important bugs fix that have been fixed in the new release (so 3.9.8 right now ?). And if you have some documentation about it, it would be nice ! Thank you in advance ! -- Mylène JOSSERAND

3 5

Kexec drops to initramfs prompt
by Paul Davies C June 29, 2013

June 29, 2013

Hi, I am trying to use the kexec to load a compiled kernel in a virtual machine that is ran using qemu. Here are the steps that I followed. 1. Compiled kernel on my Ubuntu to create bzImage. 2. Copied this bzImage to the virtual machine(debian) 3. In the virtual machine, I created an initial ramdisk using the command *mkinitramfs -o initrd.img* 4. I loaded the image to the kernel using the command *kexec -l bzImage --append="`root=UUID=<string> ro quiet`" --initrd=initrd.img*. The append was done based on /proc/cmdline. 5. Now I did a *kexec -e * which after some warning messages dropped me in to an initramfs prompt. One of the warning read as follows */dev/disk/by-uuid/<string> * *does not exist.* Can you please tell me what I am doing wrong? Please note that I have *CONFIG_KEXEC=y *on my current kernel of debian VM . -- *Regards,* *Paul Davies C* vivafoss.blogspot.com

1 0

kernel crash and dabt_svc
by Mohammed Shafi June 28, 2013

June 28, 2013

Hi, I am debugging a kernel crash in arm platform [ 486.447220] [<c000de98>] (__dabt_svc+0x38/0x60) from [<bf05c324>] (ath6kl_deliver_frames_to_nw_stack+0xac/0xf0 [ath6kl_core]) Does dabt_svc is similar to 'EIP' ? ath6kl_deliver_frames_to_nw_stack+0xac points to the exact offset (over gdb) gives the line of crash ? crash log [ 486.447220] [<c000de98>] (__dabt_svc+0x38/0x60) from [<bf05c324>] (ath6kl_deliver_frames_to_nw_stack+0xac/0xf0 [ath6kl_core]) [ 486.458520] [<bf05c324>] (ath6kl_deliver_frames_to_nw_stack+0xac/0xf0 [ath6kl_core]) from [<bf04a184>] (do_recv_completion+0x3c/0x60 [ath6kl_core]) [ 486.471692] [<bf04a184>] (do_recv_completion+0x3c/0x60 [ath6kl_core]) from [<bf04bac0>] (ath6kl_htc_pipe_rx_complete+0x5e0/0x670 [ath6kl_core]) [ 486.484563] [<bf04bac0>] (ath6kl_htc_pipe_rx_complete+0x5e0/0x670 [ath6kl_core]) from [<bf063f2c>] (ath6kl_core_rx_complete+0x10/0x14 [ath6kl_core]) [ 486.497816] [<bf063f2c>] (ath6kl_core_rx_complete+0x10/0x14 [ath6kl_core]) from [<bf08afd4>] (ath6kl_usb_io_comp_work_rx+0x34/0x7c [ath6kl_usb]) [ 486.510691] [<bf08afd4>] ( thanks in advance. regards, shafi

3 3

pci_iomap() and pci_ioremap_bar()
by Warlich, Christof June 28, 2013

June 28, 2013

Hi, can anyone explain when I should use void __iomem *regs = pci_iomap(device, barNumber, 0); and when void __iomem *regs = pci_ioremap_bar(device, barNumber); is more appropriate to map PCI memory? Furthermore, after the mapping succeeded, when may I directly access the mapped memory, e.g.: char firstByte = *regs; and when should I better use ioread8() instead?: char firstBype = ioread8(regs); All listed alternatives from above seems to work equally well for me (on x86), so I'm curious to learn what might be the difference. A guess: Does ioread8() work for both memory mapped _and_ IO mapped registers, while the direct access only works for the memory mapped case? Thanks for any clarification.

2 2

printk and kernel log
by Kevin Wilson June 27, 2013

June 27, 2013

Hi, When I add printk messages, and then try to read /var/log/messages, I see this: Jun 24 03:43:06 localhost kernel: [ 1858.148661] in my_func1 Jun 24 03:44:03 localhost kernel: [ 1915.150953] in my_func2 What are the numbers is square brackets ? like [ 1858.148661] or [ 1915.150953] ? Is there a way I can get rid of them ? Regards, Kevin

5 10

File-system driver issue
by Saket Sinha June 26, 2013

June 26, 2013

I have to port a driver from 2.6.18 to 3.8.3 Look at the below declaration of current http://lxr.free-electrons.com/source/scripts/kconfig/gconf.c#L62 Where struct menu is http://lxr.free-electrons.com/source/scripts/kconfig/expr.h#L159 The above one is in kconfig while the lower one is in the kernel. So would they be able to affect each other? I guess NO NOW in the driver to be ported- current is defined as: struct task_struct *task = current; but is still able to excess the field members of task_struct http://lxr.linux.no/linux+v2.6.18/include/linux/sched.h#L767 like attr.ia_uid = current->uid; attr.ia_gid = current->gid; though it has not been initialized or assigned or typecasted to this task_struct. HOW? I searched on the entire lxr but could not find the solution. Kindly help. PS: This driver is built on 2.6.28 kernel so use lxr from 2.6.18 kernel only like http://lxr.linux.no/linux+v2.6.18/<http://lxr.linux.no/linux+v2.6.18/include/linux/sched.h#L767>

2 4

File system query
by Dibyayan Chakraborty June 25, 2013

June 25, 2013

Hi, I came by a really weird macro code in linux/jbd.h<http://lxr.free-electrons.com/source/include/linux/jbd.h#L1029>which is BUFFER_TRACE(bh,info) do{} while(0). Please explain the necessity of this. -- With Regards Dibyayan Chakraborty

3 2

any value in "pci_request_region_exclusive()" routine?
by Robert P. J. Day June 25, 2013

June 25, 2013

poking around in memory allocation code and ran across this in drivers/pci/pci.c. first, there's this generic region request routine that accepts a final parameter as to whether it's an "exclusive" request: static int __pci_request_region(struct pci_dev *pdev, int bar, const char *res_name, int exclusive) { struct pci_devres *dr; ... and, not surprisingly, there are a couple more user-friendly routines that wrap around it: int pci_request_region(struct pci_dev *pdev, int bar, const char *res_name) { return __pci_request_region(pdev, bar, res_name, 0); } ... snip ... int pci_request_region_exclusive(struct pci_dev *pdev, int bar, const char *res_name) { return __pci_request_region(pdev, bar, res_name, IORESOURCE_EXCLUSIVE); } the above would normally suggest that, since the generic routine's name begins with "__", it's typically a lower-level routine that wouldn't be called directly, and one would expect to see calls to the wrapper routines: * pci_request_reqion() * pci_request_region_exclusive() scattered throughout the kernel source tree. there are certainly enough calls to the first (non-exclusive) routine, but this is it for the second: $ grep -rw pci_request_region_exclusive * drivers/pci/pci.c: * pci_request_region_exclusive - Reserved PCI I/O and memory resource drivers/pci/pci.c:int pci_request_region_exclusive(struct pci_dev *pdev, int bar, const char *res_name) drivers/pci/pci.c:EXPORT_SYMBOL(pci_request_region_exclusive); include/linux/pci.h:int __must_check pci_request_region_exclusive(struct pci_dev *, int, const char *); $ in short, *nobody* appears to call that second routine for exclusive access. am i missing something here? if you search for calls to the generic routine, this is all you find: $ grep -rw __pci_request_region * drivers/pci/pci.c: * __pci_request_region - Reserved PCI I/O and memory resource drivers/pci/pci.c:static int __pci_request_region(struct pci_dev *pdev, int bar, const char *res_name, drivers/pci/pci.c: return __pci_request_region(pdev, bar, res_name, 0); drivers/pci/pci.c: return __pci_request_region(pdev, bar, res_name, IORESOURCE_EXCLUSIVE); drivers/pci/pci.c: if (__pci_request_region(pdev, i, res_name, excl)) $ so what's the value of that "exclusive" wrapper routine? rday -- ======================================================================== Robert P. J. Day Ottawa, Ontario, CANADA http://crashcourse.ca Twitter: http://twitter.com/rpjday LinkedIn: http://ca.linkedin.com/in/rpjday ========================================================================

1 0

Fwd: How to calculate the TCP checksum
by Adel Qodmani June 25, 2013

June 25, 2013

On Tue, Jun 25, 2013 at 6:19 AM, Peter Teoh <htmldeveloper(a)gmail.com> wrote: > c. if your hardware is not capable to do TCP offloading, then the LONG > way is to start from beginning of packet to end of packet, and calculate > checksum. > I've tested that, my hardware apparently doesn't support TCP offloading. > d. if u keep the previous checksum, ie (th->checksum=0 is not needed > at all), and u know exactly which byte u modify, then just do a xor of the > modified byte with the previous checksum, and u get the new > checksum.....this is the preferred and fastest way. > > so in summary: > > th->check = 0; > th->check = csum_tcpudp_magic(iph->saddr, iph->daddr, > datalen, iph->protocol, > csum_partial((char *)th, datalen, 0)); > skb->ip_summed = CHECKSUM_UNNECESSARY; > > your setting of "0" above is correct, is because u want to calculate the > checksum OVERALL....ie, packet start to packet end......demarcate correctly > and u should get the answer....NOT csum_partial().....partial checksumming > does not apply here. > > -- > Regards, > Peter Teoh > Well, the problem is now solved, it seems that the error was with the fact that some skbs are not linear and I was ignoring this fact. Fixed it by the following: tcph = tcp_hdr(skb); if(skb_is_nonlinear(skb)) skb_linearize(skb) tcplen = (skb->len - (iph->ihl << 2));/* tcplen is the length of the skb - the ip-header length */ tcph->check = 0; tcph->check = tcp_v4_check(tcplen, iph->saddr, iph->daddr, csum_partial((char*) tcph, tcplen, 0)); skb->ip_summed = CHECKSUM_NONE;/* in case another hardware has TCP offloading */ Thanks a lot, anyhow :D

1 0