April 2017 - kernelnewbies@kernelnewbies.org - Kernelnewbies Mailing Lists

where is __percpu_seg initialized?
by 落痕 March 28, 2018

March 28, 2018

Hi all: I knew the kernel implement the per-cpu variable via __percpu_seg. For example on x86 architecture, it's defined as: #ifdef CONFIG_X86_64 #define __percpu_seg gs #define __percpu_mov_op movq #else #define __percpu_seg fs #define __percpu_mov_op movl #endif But I'm curious where __percpu_seg is initialized?? I fail to find the relative code. Thanks.

2 2

How to handle Hotplug with UIO userspace driver
by Mandeep Sandhu Sept. 28, 2017

Sept. 28, 2017

Hi All, We're working on a custom PCIe based hardware, which can be hotplugged into the system. We're using the UIO approach of having a minimal kernel driver which basically handles the interrupt and provides a mmap of the device memory to userspace and userspace driver which mmap's the provided memory and manages the device. We're using systemd and udev rules to spawn off a user process whenever the hardware is plugged into the system. Now my question is, how can we gracefully inform the userspace process if someone pulls out the hardware (sitting on a PCI slot). Even with full hotplug support, there can always be a case of "surprise removal", where someone yanks the board without going through the whole "attention button" routine. While the kernel driver has no problem with hot-removal, I see that if I "unregister" our UIO driver in response to the removal, then the mapping for the hardware's physical memory would become "invalid" (done by UIO on unregsiter). This could possibly cause the user-space process to crash due to illegal memory access. Is there a "standard" way to handle this scenario, i.e for hotpluggable hardware using UIO? Basically, we'd want to "unregister" the UIO device only after the userspace process is informed of the removal so that it could unmap/close the device and exit gracefully. We could try doing it using some systemd trick, where we put a dependency on one of the UIO created files (/dev/uioX), so that when the file is destroyed, systemd could kill/shutdown the process. Although, there's no guarantee that the userspace process won't try to access the mmaped space during that interval, so this isn't a reliable solution either. Any hints, thoughts? Thanks., -mandeep

3 7

devm_ioremap() in a sysfs show method
by Martin Kaiser May 9, 2017

May 9, 2017

Dear all, I'm trying to create a simple sysfs node that prints the content of a register. The code looks like this static ssize_t my_read_reg(struct device *pdev, struct device_attribute *attr, char *buf) { u32 __iomem *reg; reg = devm_ioremap(pdev, 0x581018C4, 4); return scnprintf(buf, PAGE_SIZE, "0x%02x\n", ioread8(reg)); } Is it ok to use devm_ioremap() on the device that I get here? If so, may I assume that devm_ioremap() re-uses an existing mapping and it's ok to call it for each sysfs access? The physical address that I pass to devm_ioremap() is not page-aligned. This seems to work on arm. Is there a recommended (i.e. more portable) way of doing this? Should I map a larger area and use PAGE_ALIGN()? Thanks for your feedback, Martin

1 1

endian patches
by Tobin C. Harding May 1, 2017

May 1, 2017

Should [drivers/staging/*] patches to endian code be tested on hardware before submission? During recent development of ks7010 driver, and from watching patch review on devel(a)linuxdriverproject.org, I formed the opinion that patches fixing endian issues need to be tested on hardware before they can be *guaranteed* to be correct. Is this opinion well founded? thanks, Tobin.

3 3

Other than modifying ICSK_CA_PRIV_SIZE, any way to add more info to a tcp socket?
by Shiyao Ma April 30, 2017

April 30, 2017

Hi, When defining my own TCP congestion control, I encountered this: BUILD_BUG_ON(sizeof(struct mytcp) > ICSK_CA_PRIV_SIZE); Every socket using my own congestion control has an instance of the struct mytcp, which stores per-socket information. The above bug means the size of struct mytcp exceeds the limit of kernel's predefined per-socket struct size. Provided the struct mytcp cannot be trimmed down any more, and meanwhile ICSK_CA_PRIV_SIZE can not be increased, is there any way to add more info to a socket ? I am thinking about making a certain filed of the struct mytcp be a pointer, so that when the socket gets created, a hook function is invoked to kmalloc some memory to store extra information. And when the socket is destroyed, that memory is also destroyed by some hook function. But this idea seems a little bit messy, and I dunno what the exact hook functions are. Any other remedies? Best,

1 0

Need help with tracepoints
by Matwey V. Kornilov April 29, 2017

April 29, 2017

Hello, I am following trace/tracepoints.txt guide to add some tracepoints to already existing kernel module. I've created include/trace/events/mysubsys.h and added #include <trace/events/mysubsys.h> #define CREATE_TRACE_POINTS DEFINE_TRACE(mysubsys_event1); EXPORT_TRACEPOINT_SYMBOL_GPL(mysubsys_event1); to the module sources. There are symbols *_mysubsys_event1 in module ko file. The issue is that I don't see my events in /sys/kernel/debug/tracing/events Not even ./mysubsys path here.

2 3

kgdb debug with ttyUSB0 instead of ttyS0
by Oliver Graute April 29, 2017

April 29, 2017

Hello list, i'am using a null modem cable with two usb to serial converts on both ends between my develop and my target machine. The connection is fine. On both machines the serial interface is /dev/ttyUSB0 for this connection. On the target machine I try to enable kgdboc echo ttyUSB0 > /sys/module/kgdboc/parameters/kgdboc but only got: echo write error: No such device if I try: echo ttyS0 > /sys/module/kgdboc/parameters/kgdboc is not complaining. Some ideas why ttyUSB0 can't be used with kgdb? Best regards, Oliver

4 6

question on patch submission
by Michael Mera April 28, 2017

April 28, 2017

Hello, As part of the eudyptula challenge I submitted a coding style patch on staging (that you can find here: https://lkml.org/lkml/2017/4/28/148) I had read beforehand the submission guidelines, but I still have some questions which I think are more appropriate posted here than as an answer to the patch thread: - the patch is clear with "checkpatch.pl --strict", but there are some tabs in the middle of the line. In my editor and apparently with checkpatch.pl those tabs are used for alignment (so spend less than 8 characters) so that the total of the line in less than 80 characters. In the submitted patch however, due to the +- at the beginning of the line they appear larger and thus the line in more than 80 character. My question is: Should I consider the patch clean or interpret the guideline "tabs are 8 characters" strictly (and not tabs "at the beginning on the line" are 8 characters)? - when I make modification to the patch message but not the diff lines, should I still increment the patch version, or rather maybe resend it with something like RESEND in the subject? Sorry if these are written in the doc, I reread it but couldn't find clear answers (and sorry to Greg Kroah-Hartma who has to read my dumb submissions both here and in the kernel mailing list). Thank you, Michael Mera

2 2

Task in Linux kernel
by Nagaraju Bellapu April 27, 2017

April 27, 2017

Hi All, What is the task put to sleep and schedule term referring while using wait_queue_event and wake_up functions in a kernel module code.? Please help me in understanding the wait_queues in linux kernel. regards nagaraju

2 1

english grammar and patches
by Tobin C. Harding April 26, 2017

April 26, 2017

This question relates to English grammar and correct usage when writing gitlog messages and patch series cover letters. The writing of gitlog messages is covered in submitting-patches.rst, of note is the mood to use. It is not stated but I think it is a subjunctive description of the problem being addressed followed by an imperative description of what is being done in the patch. Please correct me if I am wrong. The question is: what mood to use in the cover letter. Also, is the subject of the cover letter explicit or implicit? Said another way is it acceptable to say 'this series' within the cover letter? It is specifically stated not to use the subject within the patch itself, however, many patch sets on LKML use the subject in the cover letter. Finally, has anyone, within easy reach, a link to a canonical patch set cover letter. I have searched LKML but am unable to judge for myself. I would like to add a link to the kernel newbies wiki also if one is available. thanks, Tobin.

3 3