July 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

Re: new sysctl tunable knob for tcpip
by Massimo Sala Aug. 1, 2017

Aug. 1, 2017

I wish to suggest to developers to add this new knob : tcp_retries2_time - INTEGER This value influences the timeout (seconds) of an alive TCP connection, when RTO retransmissions remain unacknowledged. RFC 1122 recommends at least 100 seconds for the timeout, which corresponds to a value of at least 8. Rationale : * I see too many abuse of tcp_retries2, because even if you check the RTO, it will change with different connections, routing, etc. * It will be friendly and I think better to document an absolute value. See other parameters with two knobs ( example : dirty_ratio, dirty_ratio_bytes ). >> q1) which is the right kernel m-list to discuss it ? > net...@vger.kernel.org > Good luck! > greg k-h mlist : netdev(a)vger.kernel.org ? Best regards, Sala

2 2

new sysctl tunable knob for tcpip : errata corrige
by Massimo Sala July 31, 2017

July 31, 2017

> net...@vger.kernel.org > greg k-h I see now the emails are trunated. Greg, I presume you wrote "netdev" ? Sala ---------- Forwarded message ---------- From: Massimo Sala <massimo.sala.71(a)gmail.com> Date: Mon, 31 Jul 2017 15:16:34 +0200 Subject: Re: new sysctl tunable knob for tcpip To: kernelnewbies(a)kernelnewbies.org I wish to suggest to developers to add this new knob : tcp_retries2_time - INTEGER This value influences the timeout (seconds) of an alive TCP connection, when RTO retransmissions remain unacknowledged. RFC 1122 recommends at least 100 seconds for the timeout, which corresponds to a value of at least 8. Rationale : * I see too many abuse of tcp_retries2, because even if you check the RTO, it will change with different connections, routing, etc. * It will be friendly and I think better to document an absolute value. See other parameters with two knobs ( example : dirty_ratio, dirty_ratio_bytes ). >> q1) which is the right kernel m-list to discuss it ? > net...@vger.kernel.org > Good luck! > greg k-h mlist : netdev(a)vger.kernel.org ? Best regards, Sala

1 0

OOM killer hung the whole system
by Rock Lee July 31, 2017

July 31, 2017

Hi, I was using a program to test OOM killer, but OOM killer hung my board. The test snippet is very simple, just malloc() without free(). 6 int i=0; 7 char *ptrtest=NULL; 8 while (1) 9 { 10 ptrtest = malloc(0x1000); 11 if (ptrtest == NULL) 12 printf("malloc failed.\n"); 13 } 14 return 0; 15 After I executed this program(a.out), there is no reaction of my shell. Only this messege was shown: [ 180.138188] Out of memory: Kill process 2706 (qq) score 623 or sacrifice child [ 180.164230] Killed process 2706 (qq) total-vm:98156kB, anon-rss:96400kB, file-rss:4kB I think it must be some problem with OOM killer, and tried a lot of way to get some infomation. And I got this: ======================================================= Process: a.out, cpu: 0 pid: 1463 start: 0xd5849b00 ===================================================== Task name: a.out pid: 1463 cpu: 0 state: 0x2 exit_state: 0x0 stack base: 0xd6190000 Stack: [<c08394e4>] __schedule+0x2c8 [<c018df88>] squashfs_cache_get+0x108 [<c0191020>] squashfs_readpage_block+0x28 [<c018f648>] squashfs_readpage+0x624 [<c0090d4c>] __do_page_cache_readahead+0x228 [<c00882f4>] filemap_fault+0x1d4 [<c00a5f10>] __do_fault+0x34 [<c00a8ab0>] do_read_fault+0x19c [<c00a9388>] handle_mm_fault+0x468 [<c00164b0>] do_page_fault+0x11c [<c00084a4>] do_PrefetchAbort+0x34 [<c0011c1c>] ret_from_exception+0x0 I am suprised, because a.out is in /tmp, there is no relation between a.out and squashfs(my rootfs). Could anyone give me a hint ? Is the problem of squashfs or OOM killer. -- Cheers, Rock

4 5

Re: OOM killer hung the whole system
by Rock Lee July 31, 2017

July 31, 2017

On Mon, Jul 31, 2017 at 3:51 AM, Mulyadi Santosa <mulyadi.santosa(a)gmail.com> wrote: > > > On Tue, Jul 25, 2017 at 4:59 PM, Rock Lee <rockdotlee(a)gmail.com> wrote: > >> Hi, >> >> I was using a program to test OOM killer, but OOM killer hung my board. >> The test snippet is very simple, just malloc() without free(). >> >> 6 int i=0; >> 7 char *ptrtest=NULL; >> 8 while (1) >> 9 { >> 10 ptrtest = malloc(0x1000); >> 11 if (ptrtest == NULL) >> 12 printf("malloc failed.\n"); >> 13 } >> 14 return 0; >> 15 >> >> After I executed this program(a.out), there is no reaction of my shell. >> Only this messege was shown: >> >> [ 180.138188] Out of memory: Kill process 2706 (qq) score 623 or >> sacrifice child >> [ 180.164230] Killed process 2706 (qq) total-vm:98156kB, >> anon-rss:96400kB, file-rss:4kB >> >> > I think the clue is above. Could you find out what is the process with pid > 2706? > > I bet that is the shell that runs as the parent of your a.out program, but > feel free to make sure what that is. > Sorry, my bad, it is a spelling mistake. a.out is actually qq (pid 2706) > > > >> I think it must be some problem with OOM killer, and tried a lot of way >> to get some infomation. And I got this: >> >> ======================================================= >> Process: a.out, cpu: 0 pid: 1463 start: 0xd5849b00 >> ===================================================== >> Task name: a.out pid: 1463 cpu: 0 >> state: 0x2 exit_state: 0x0 stack base: 0xd6190000 >> Stack: >> [<c08394e4>] __schedule+0x2c8 >> [<c018df88>] squashfs_cache_get+0x108 >> [<c0191020>] squashfs_readpage_block+0x28 >> [<c018f648>] squashfs_readpage+0x624 >> [<c0090d4c>] __do_page_cache_readahead+0x228 >> [<c00882f4>] filemap_fault+0x1d4 >> [<c00a5f10>] __do_fault+0x34 >> [<c00a8ab0>] do_read_fault+0x19c >> [<c00a9388>] handle_mm_fault+0x468 >> [<c00164b0>] do_page_fault+0x11c >> [<c00084a4>] do_PrefetchAbort+0x34 >> [<c0011c1c>] ret_from_exception+0x0 >> >> I am suprised, because a.out is in /tmp, there is no relation between >> a.out and squashfs(my rootfs). Could anyone give me a hint ? Is the problem >> of squashfs or OOM killer. >> > > The above stack trace just mention what happened in kernel space. Very > likely when it was about to be killled by OOM killer. CMIIW. > > Thus, it has little or nothing to do with the fact that it called squash > fs related function > There is a chance squashfs functions will be called when out of memory. It is when a libary file is mmaped, filemap_fault() gave me a hint. > > > > -- > regards, > > Mulyadi Santosa > Freelance Linux trainer and consultant > > blog: the-hydra.blogspot.com > training: mulyaditraining.blogspot.com > > > <https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campai…> Virus-free. > www.avast.com > <https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campai…> > <#m_-5780323552132534074_DAB4FAD8-2DD7-40BB-A1B8-4E2AA1F9FDF2> > > _______________________________________________ > Kernelnewbies mailing list > Kernelnewbies(a)kernelnewbies.org > https://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies > > -- Cheers, Rock

1 0

GRUB always enters a rescue shell.
by Hao Lee July 27, 2017

July 27, 2017

Hi, I encounter a problem when making a bootable disk image because GRUB always enters a rescue shell. I want to create a disk image and install GRUB 2 on it and then use Bochs to run it. The steps are as follows. System Environment: OS: Fedora 25 x86 GRUB version: grub2-install (GRUB) 2.02~beta3 Bochs version: 2.6.9 1. Create a empty disk image. dd if=/dev/zero of=hd.img bs=$((16*63*512)) count=100 2. Create a primary partition. fdisk -u -C100 -S63 -H16 hd.img The partition table is: Device Boot Start End Sectors Size Id Type hd.img1 2048 100799 98752 48.2M 83 Linux 3. Attach the whole image to /dev/loop0 and attach the first partition to /dev/loop1 sudo losetup /dev/loop0 hd.img sudo losetup -o $((2048*512)) /dev/loop1 hd.img 4. Format the first partition to ext4 and mount it to /mnt/ sudo mkfs.ext4 /dev/loop1 sudo mount -t ext4 /dev/loop1 /mnt 5. Create GRUB directory. sudo mkdir -p /mnt/boot/grub2 6. Install GRUB 2 sudo grub2-install --boot-directory=/mnt/boot /dev/loop0 7. Start Bochs. bochs -f test.bxrc These steps are straight and simple. However, GRUB always enters a rescue shell instead of entering normal mode. When I use the set and ls command to show more information, I can only see (hd0) and the first partition (hd0,1) is disappeared. The screen output is as follows: ---------------------------------------------------- Bochs BIOS - build: 02/16/17 ata0 master: Generic 1234 ATA-6 Hard-Disk ( 49 MBytes) Press F12 for boot menu. Booting from Hard Disk... . error: no such device: c5b2c483-70c2-4212-840b-da5f965d0555. error: unknown filesystem. Entering rescue mode... grub rescue> set prefix=(hd0)/boot/grub2 root=hd0 grub rescue> ls (hd0) grub rescue> ---------------------------------------------------- Could someone help me out? Many thanks! Regards, Hao Lee

3 6

Compiling customer kernel for Debian 9.X
by HP Garcia July 26, 2017

July 26, 2017

What's to the procedure for compiling a custom kernel for Debian 9.X. Thanks H.P. Garcia, Photographer HPG Photography

1 0

Need help to understand Linux kernel source versioning systems
by vibnwis July 26, 2017

July 26, 2017

Hi there, Being a newbie, I encountered the following scenarios when tracing build_kernel.sh script from eewiki. Would someone help me to understand why please? From my experience, I would to summarize the following scenario. 1. First, source is cloned from https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git and checked for release version, "v4.10.17". Obviously it has no kernel tag of "v4.10.17", but "v4.10" 2. As "v4.10.17" does not exist, it tried to pull from the origin/master by command, git pull --no-edit https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git master --tags 3. Kernel tag, of "v4.10.17" is checked being verified again, but failed. Hence, now it fetched from git fetch https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable.git master --tags Now, the kernel, of v4.10.17" is found. My questions are: 1. One would know that clone of https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git would not produce kernel tag of , "v4.10.17", wouldn't it? 2. Similar by pull it again from https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git with option of master --tags would not produce kernel tag, "v4.10.17" as well, wouldn' it? 3. Since fetching with git fetch https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable.git master --tags would produce v4.10.17, why would not it clone it from .../stable/linux-stable,git? I know that the answers would be very lengthy and I would probably need some basics of LInux kernel to understand the answers. However, I will investigate further for things that I don't understand. For now, please educate me if you could. The following are the snapshot of the process getting the kernel source. > git clone https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git /home/qws/myGithub/Panda-Repo/armv7-multiplatform/ignore/linux-src 2. > git clone --shared /home/qws/myGithub/Panda-Repo/armv7-multiplatform/ignore/linux-src /home/qws/myGithub/Panda-Repo/armv7-multiplatform/KERNEL 3. > git tag | grep "v4.10.17" | grep -v rc > /dev/null 2>&1 || git_kernel_torvalds # The following is resulted of git_kernel_torvalds being executed, as the former command failed > git pull --no-edit https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git master --tags 4. > git tag | grep "v4.10.17" | grep -v rc > /dev/null 2>&1 || git_kernel_stable # The following is resulted of git_kernel_stable being execuated, as the former command failed > git fetch https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable.git master --tags Thank you Best regards Lim

4 4

Contribute
by nunojsa July 24, 2017

July 24, 2017

Hi all, Im fairly new to linux kernel and would like to start with some staging work (seems like the best way to start). Basically what i would like to know is which tree should i clone for this? My understanding is that one should never work directly on top of the mainline tree, instead (for this case) on top of the linux staging tree. The same goes, for example, for USB subsystem development in which case i should just clone the usb development tree (and the same for all subsystems). Is my understanding correct? Thank you in advance!

3 4