- kernelnewbies@kernelnewbies.org - Kernelnewbies Mailing Lists

How to disable address randomization ?
by admin LI Jan. 14, 2022

Jan. 14, 2022

Hi, I'm developing a kernel module for an ARM machine, while debugging I found addresses printed are all randomized and useless for debugging. To prove I was not crazy I wrote this small program: --------------------------------- #include <linux/init.h> #include <linux/module.h> #include <linux/kernel.h> #include <linux/slab.h> MODULE_LICENSE("GPL"); MODULE_AUTHOR("Somebody"); MODULE_DESCRIPTION("A simple example Linux module."); MODULE_VERSION("0.01"); static int __init example_init(void) { uint32_t a; uint32_t b; uint32_t c; uint8_t d[10]; uint8_t *e; printk(KERN_INFO "Hello, World!\n"); printk(KERN_INFO "&a %p\n",&a); printk(KERN_INFO "&b %p\n",&b); printk(KERN_INFO "&c %p\n",&c); printk(KERN_INFO "&d %p\n",d); printk(KERN_INFO "&d[0] %p\n",&d[0]); printk(KERN_INFO "&d[1] %p\n",&d[1]); e = kmalloc(10, GFP_KERNEL); printk(KERN_INFO "&e[0] %p\n",&e[0]); printk(KERN_INFO "&e[1] %p\n",&e[1]); kfree(e); return 0; } static void __exit example_exit(void) { printk(KERN_INFO "Goodbye, World!\n"); } module_init(example_init); module_exit(example_exit); --------------------------------- And it gave me this output: Hello, World! &a b3f9fa31 &b 27e1c68a &c da50d287 &d 9f9aec2b &d[0] 9f9aec2b &d[1] cc627580 &e[0] 98b8c9eb &e[1] 45f248f8 Then I tested on my debian host machine which gave me the same kind of randomized addresses. When I search randomization the only thing I found is KASLR which I don't think is the same thing.

5 7

Does regmap-i2c not automatically switch to I2C mux?
by www Jan. 12, 2022

Jan. 12, 2022

Dear all， Does regmap-i2c not automatically switch to I2C mux? Add i2c-device(the driver use regmap-i2c) under i2c-mux on a i2c bus, set i2c-mux to another channel, then cat the node under /sys/class/hwmon/xxx, cannot get the value from the i2c-device. when use another i2c-device( not use regmap in driver) get value from i2c-device, this error will not occur. Does the device driver need any special settings when using regmap? Does anyone know why? thanks, Byron

1 0

Documentation confusion on ICMP Rate Mask/Rate Limiting
by Daryll Swer Jan. 9, 2022

Jan. 9, 2022

Hi Folks So I will get straight to the point, I am trying to figure out some answers on *icmp_ratelimit* and *icmp_ratemask* based on the man page <https://mailtrack.io/trace/link/1ab83c495d25602c6b2de3095e1a1fa4b795d988?ur…> . The questions are: 1. How can we determine the Packet per second rate on any given rate-limit value (say 10)? Without of course going the hard way of benchmarking it by ICMP flooding. 2. Just what exactly is the *correct mask* in *Binary form* to *include all* known ICMP types instead of just the default mask? Been having a hard time with this one. 3. Do we have something of this nature for *IPv6 *in the Kernel? I don't see it in most Linux based NetworkOSes as a documented feature. Default rate-mask below, just can't get it through my head. Maybe we should also look at revising the documentation wordings for it and possibly make some changes to the actual code to simplify the bit mapping? My best guess is that is is the answer to question *2*: *1111111100100111001* [image: image.png] *--* Best Regards Daryll Swer Website: daryllswer.com <https://mailtrack.io/trace/link/2f87aeb108f5e5520487e1f3abc5bad533b8b292?ur…>

2 2

Customize path to kernel deb
by Lafan Mining Jan. 9, 2022

Jan. 9, 2022

I have a kernel sources at /home/lafan/linux. After compiling the kernel with make make -j $(getconf _NPROCESSORS_ONLN) deb-pkg LOCALVERSION=-custom The resulting deb package is created at /home/lafan. Is there a way to customize where it is created?

2 1

x86, nmi: IOCK error
by Muni Sekhar Jan. 8, 2022

Jan. 8, 2022

Hi all, The following message is seen on the console "NMI: IOCK error (debug interrupt?) for reason 60 on CPU 0.", what does it mean? Does this likely confirm the CPU is faulty? If so, how to resolve this? -- Thanks, Sekhar

3 3

Contiguous Memory Allocator (CMA)
by Sadanand Warrier Jan. 8, 2022

Jan. 8, 2022

Hello Is CMA still supported as described in this document https://lwn.net/Articles/396707/ ? Or has it changed and there is a newer version which details how to allocate a chunk of contiguous memory on bootup and then use it with a loadable kernel module? What I want to do is allocate a chunk of contiguous memory at boot time and use it with a loadable driver. It will be a vanilla Centos kernel and I wonder if they are compiled with the CMA enabled. When I check dmesg I see " 0K cma-reserved" so it still exists. Thanks S

2 4

Device to test usblcd.c wanted
by Philipp Hortmann Dec. 22, 2021

Dec. 22, 2021

Hi all, I would like to test drivers/usb/misc/usblcd.c with a USB LCD Display. But I cannot find a suitable device. I have searched with google but I get so much hits for other devices that I cannot find the right one. I have tried to contact the author of the driver but: The response from the remote server was: 550 5.1.1 <g.toth(a)e-biz.lu>: Recipient address rejected: User unknown The idVendor = 0x10D2 does also not help because RayComposer - R. Adams does not response and does not offer such products. It seems that in 2009 the idVendor = 0x10D2 was used by vendor MOLEX. But this also does not help. Thanks for your support. Bye Philipp

1 0

[IMA-EVM] When "appraise" applied, kernel cannot start init process.
by Kim, Jeong-Hwan Dec. 19, 2021

Dec. 19, 2021

Hi, everyone Now I'm testing ima-evm security module to see if it can be applied my embedded system board. (not using TPM) The kernel version is 5.4.91 (based on beaglebone black board). But, unfortunately I cannot verify the functions of ima-evm correctly. When I apply "appraise", kernel panic happens. This is what happened: (1) I set ima-evm keys as follows : # dd if=/dev/urandom bs=1 count=32 status=none | keyctl padd user kmk-user @u # keyctl link @u @s #keyctl pipe `key search @u user kmk-user` > /etc/keys/kmk-user.blob #keycl add encrypted evm-key "new user:kmk-user 32" @u #keyctl pipe `keyctl search @u encrypted evm-key` > /etc/keys/evm-user.blob # openssl genrsa -out /etc/keys/rsa_private.pem 1024 # openssl rsa -pubout -in /etc/keys/rsa_private.pem -out /etc/keys/rsa_public.pem (2) I set kernel parameters with *"ima_policy=appraise_tcb ima_appraise=fix evm=fix"* (3) After boot, I loaded the keys lik this: # keyctl add user kmk-user "`cat /etc/keys/kmk-user.blob`" @u # keyctl link @u @s # keyctl add encrypted evm-key "load `cat /etc/keys/evm-user.blob`" @u # evmctl import --rsa /etc/keys/rsa_public.pem $(keyctl newring _ima @u) # evmctl import --rsa /etc/keys/rsa_public.pem $(keyctl newring _evm @u) # echo 1 > /sys/kernel/security/evm (4) I created the ima-evm digital signatures : # find / -type f -exec evmctl ima_sign --key /etc/keys/rsa_private.pem '{}' \; (5) I changed the kernel parameter as this : *"ima_policy=appraise_tcb ima_appraise=enforce"* (6) When the board reboot, it stop because of kernel panic : *[ 3.546536] audit: type=1800 audit(946688023.636:2): pid=1 uid=0 auid=4294967295 ses=4294967295 subj==unconfined op=appraise_data cause=unknown comm="swapper" name="/lib/systemd/systemd" dev="mmcblk0p1" ino=5684 res=0* When the kernel parameter is "ima_policy=tcb ima_appraise=enforce", kernel panic does not happen, but after boot, any executable file is not blocked although its contents is changed. Please give me an advice. Thanks in advance, J.Hwan Kim

1 0

X86 : system call
by Muni Sekhar Dec. 12, 2021

Dec. 12, 2021

Hi all, 32-bit system call numbers and entry vectors are defined in the format "<number> <abi> <name> <entry point> <compat entry point>". What is the difference between <entry point> and <compat entry point>? The abi is always "i386" for arch/x86/entry/syscalls/syscall_32.tbl, whereas for 64-bit(arch/x86/entry/syscalls/syscall_64.tbl) it is used as "common" and "64", "x32". What is the difference between "common" and "64", "x32"? While adding a new system call(for example: mlock) to kernel, /kernel/sys_ni.c file modified as mentioned below: +cond_syscall(sys_mlock2); Why do we pass <entry point> to cond_syscall()? Also, could someone please point me to a document which explains about "how to add a new system call" and the "system call flow from user space to kernel space" in X86 ARCHITECTURE (32-BIT AND 64-BIT). -- Thanks, Sekhar

2 1

Time: new clocksource
by Muni Sekhar Dec. 8, 2021

Dec. 8, 2021

Hi All, We have a Digital PLL with 64 bit timer counter hardware and the counter is accessible from the CPU over the PCIe bus. Is it possible to add this timer counter hardware as new clocksource driver? To do this, can someone please point me to the existing reference source code(or patch) for this task. Suppose if it is possible to add a new clocksource driver for this hardware then does any userspace get_timestamp* API would get the time from this new hardware? -- Thanks, Sekhar

3 6