July 2016 - kernelnewbies@kernelnewbies.org - Kernelnewbies Mailing Lists

gpio-led driver
by Raul Piper July 21, 2016

July 21, 2016

Hi, I wanted to know the part number for the leds-gpio.c in the driver/leds folder and the device tree bindings for this driver .Can some one point out to me where in Linux kernel it is? Also why below line has been used.Is it mandatory? ... *.of_match_table = of_gpio_leds_match,* }, thanks in advance , Rp

3 3

Mapping of Filename to allocated blocks
by Manoj Rao July 21, 2016

July 21, 2016

Hi All, I'm looking for a way to get a filename associated with a given physical block (ideally what I'd like is a mapping of filename <=> all the allocated blocks for this file). Is there a recommended way to do this already in kernel? if not, then where should I start looking to add changes? Help appreciated. Thanks.

3 3

How do you test Linux kernel?
by Aleksander Alekseev July 20, 2016

July 20, 2016

Hello It is my understanding that Linux kernel doesn't has usual `make check` or `make test` command. There are LTP [1] and other related projects instead. Could you tell me please what test suits (if any) do _you_ run in practice? [1] http://linux-test-project.github.io/ -- Best regards, Aleksander Alekseev

3 2

How old Linux Device Drivers 3 book is now?
by Fernando França July 20, 2016

July 20, 2016

Hi everyone. I'm a totally fresh and newbie with kernel development and I've started reading Linux Device Drivers 3 book (CORBET et al). Since this book approaches is based on kernel 2.6 and we are currently with kernel 4.x, someone could help me telling if this book is still a good introduction to the subject, or is now highly outdated? Thanks in advance. -- Fernando França http://desconstruindo.eng.br Linux user #263682 Dificultando a vida dos usuários desde 1999

2 1

kgdb on kernel4.4 for arm
by Pankaj Gupta July 20, 2016

July 20, 2016

Hello all, I am trying to use kgdboc on arm64 based kernel 4.4 android device, but could not able to connect it. Is anyone have used kgdb on arm based android devices? Thanks & Regards, *Pankaj Gupta*

1 0

Pending work related to kernel for a newbie
by Ashijeet Acharya July 20, 2016

July 20, 2016

Hi everyone, I am kernel newbie and I was looking for something to work upon to improve my skill-set. I have good knowledge of C and experience in writing patches. Also please refer me if there is a to-do list for newbies related to pending tasks for linux kernel. Thanks Ashijeet

4 3

Interrutp from USB I/O Expander.
by Daniel. July 19, 2016

July 19, 2016

Hi everybody. I know that there are some GPIO lines that needs sleeping to get actived/deactived, this is why we have the gpio_cansleep_*() functions. So, I'm planing using the Microchip MCP2210 which act as USB->SPI bridge. The problem is that the SPI slave device has two extra lines, one for IRQ and another for enabling the chip (CE). Since MCP2210 has GPIOs too I was wondering about wiring IRQ and CE to some GPIO lines and using it as general GPIO at host. My concern is if the IRQ line through USB, if it would sleep, if would be possible to use it as IRQ!? The summarized setup is this HOST -> usb bus -> MCP2210 <-- SPI bridge <- slave device spi lines '- GPIO 0 <- slave device IRQ line '- GPIO 1 <- slave device CE line Any help is appreciated, Best regards and thanks in advance! -- "Do or do not. There is no try" Yoda Master

1 0

drivers: staging: most: Locking question
by mhornung.linux＠gmail.com July 18, 2016

July 18, 2016

Hello, I have some questions about the locking techniques used inside file drivers/staging/most/hdm-usb/hdm_usb.c. The one and only call to function free_anchored_buffers is locked by a Mutex: ---------------------------------------------------------------------------- ... mutex_lock(&mdev->io_mutex); free_anchored_buffers(mdev, channel); if (mdev->padding_active[channel]) mdev->padding_active[channel] = false; if (mdev->conf[channel].data_type == MOST_CH_ASYNC) { del_timer_sync(&mdev->link_stat_timer); cancel_work_sync(&mdev->poll_work_obj); } mutex_unlock(&mdev->io_mutex); ... ---------------------------------------------------------------------------- Then, inside function free_anchored_buffers, they use a (from my point of view) somewhat complex spinlock variant: ---------------------------------------------------------------------------- static void free_anchored_buffers(struct most_dev *mdev, unsigned int channel) { struct mbo *mbo; struct buf_anchor *anchor, *tmp; unsigned long flags; spin_lock_irqsave(&mdev->anchor_list_lock[channel], flags); list_for_each_entry_safe(anchor, tmp, &mdev->anchor_list[channel], list) { struct urb *urb = anchor->urb; spin_unlock_irqrestore(&mdev->anchor_list_lock[channel], flags); if (likely(urb)) { mbo = urb->context; if (!irqs_disabled()) { usb_kill_urb(urb); } else { usb_unlink_urb(urb); wait_for_completion(&anchor->urb_compl); } if ((mbo) && (mbo->complete)) { mbo->status = MBO_E_CLOSE; mbo->processed_length = 0; mbo->complete(mbo); } usb_free_urb(urb); } spin_lock_irqsave(&mdev->anchor_list_lock[channel], flags); list_del(&anchor->list); cancel_work_sync(&anchor->clear_work_obj); kfree(anchor); } spin_unlock_irqrestore(&mdev->anchor_list_lock[channel], flags); } ---------------------------------------------------------------------------- To my questions: #1: What is the intention of locking a whole function with a Mutex and then using spinlocks inside the function? Wouldn't it be sufficient to use one locking technique? #2: Why is the spinlock not just locking the whole list_for_each_entry part or just the list_del(&anchor->list)? Thank you very much in advance. With best regards Michael

2 1

pushing other commits upstream
by Alexander Couzens July 18, 2016

July 18, 2016

Hi, I would like to push patches from somebody else upstream. I'm talking of commits of the TI's beagleboard repo [1]. What do I need to take care of? Best, lynxis [1] https://github.com/beagleboard/linux -- Alexander Couzens mail: lynxis(a)fe80.eu jabber: lynxis(a)fe80.eu mobile: +4915123277221 gpg: 390D CF78 8BF9 AA50 4F8F F1E2 C29E 9DA6 A0DF 8604

4 3

Different ways of doing a system call
by Aleksander Alekseev July 17, 2016

July 17, 2016

Hello Recently I've explored different ways of doing a system call on Linux (only on x86 and x64 architectures). Here is what I discovered so far: * int 0x80 (x86, slow) * sysenter (x86, fast) * vsyscall (x86, deprecated) * VDSO trampoline (x86, fast) * syscall (x64, fast) Tell me, please: * Didn't I miss anything? * What are _recommended_ ways of doing a system call on x86 and x64? * I just wanted to clarify - in theory it's OK to use both say int 0x80 and syscall in the same ELF64 binary? Wouldn't it cause a program to crash or something? -- Best regards, Aleksander Alekseev

1 0