February 2015 - kernelnewbies@kernelnewbies.org

overview of bust_spinlocks
by manty kuma Feb. 20, 2015

Feb. 20, 2015

What is the concept of bust_spinlocks() In do_kernel_fault() function of ARM 64, we can see the following code https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/tree/arch/a… static void __do_kernel_fault(struct mm_struct *mm, unsigned long addr, unsigned int esr, struct pt_regs *regs){ /* * Are we prepared to handle this kernel fault? */ if (fixup_exception(regs)) return; /* * No handler, we'll have to terminate things with extreme prejudice. */* bust_spinlocks(1);* pr_alert("Unable to handle kernel %s at virtual address %08lx\n", (addr < PAGE_SIZE) ? "NULL pointer dereference" : "paging request", addr); show_pte(mm, addr); die("Oops", regs, esr); bust_spinlocks(0); do_exit(SIGKILL);} I have looked into its definition. It is as follows: void __attribute__((weak)) bust_spinlocks(int yes){ if (yes) { ++oops_in_progress; } else {#ifdef CONFIG_VT unblank_screen();#endif console_unblank(); if (--oops_in_progress == 0) wake_up_klogd(); }} So basically when i am saying bust_spinlocks(1), it is just incrementing the oops)_in_progress variable. How is it serving the purposr of a lock here? Best Regards, Manty

1 0

sk_wait_data
by Grzegorz Dwornicki Feb. 20, 2015

Feb. 20, 2015

Hello Proces have a established tcp socket. It calls sys_revcmsg on that socket. That function goes all the way to tcp_recvmsg function. Function tcp_recvmsg.c (defined in net/ipv4/tcp.c) blocks it self on function sk_wait_data if socket recv queue is empty. I wanted to know how socked is checked in time. Soo i Looked up this sk_wait_data function (http://lxr.free-electrons.com/source/net/core/sock.c#L1933) 1933 int sk_wait_data(struct sock *sk, long *timeo) 1934 { 1935 int rc; 1936 DEFINE_WAIT(wait); 1937 1938 prepare_to_wait(sk_sleep(sk), &wait, TASK_INTERRUPTIBLE); 1939 set_bit(SOCK_ASYNC_WAITDATA, &sk->sk_socket->flags); 1940 rc = sk_wait_event(sk, timeo, !skb_queue_empty(&sk->sk_receive_queue)); 1941 clear_bit(SOCK_ASYNC_WAITDATA, &sk->sk_socket->flags); 1942 finish_wait(sk_sleep(sk), &wait); 1943 return rc; 1944 } This function blocks it self on the sk_wait_event macro. Here is it definition (http://lxr.free-electrons.com/source/include/net/sock.h#L893) 893 #define sk_wait_event(__sk, __timeo, __condition) \ 894 ({ int __rc; \ 895 release_sock(__sk); \ 896 __rc = __condition; \ 897 if (!__rc) { \ 898 *(__timeo) = schedule_timeout(*(__timeo)); \ 899 } \ 900 lock_sock(__sk); \ 901 __rc = __condition; \ 902 __rc; \ 903 }) This macro is blocked in the schedule_timeout all the time. I know this because I added printk functions with some marks easly to target the line blocking the socket. It hangs on this schedule_timeout and it magicaly released when some data apear in the revc queue. How does the kernel know when to "unblock" the process? Function name: sk_wait_data and macro name: sk_wait_event tells me that there should be some queue checning from time to time... Can someone explain this to me?

3 2

Re: Linux do_coredump() and SMP systems
by Sudharsan Vijayaraghavan Feb. 19, 2015

Feb. 19, 2015

We are doing prototype so much change have gone into kernel , we are finding it difficult to upgrade to latest immediately However I ran through the code once again, indeed kernel handles it down_write(&mm->mmap_sem); in coredump_wait() makes sure the second coredump is stopped and returns negative for core_waiters I will debug further, thanks for confirming that kernel handles this scenario On Tue, Feb 17, 2015 at 9:57 PM, Sudharsan Vijayaraghavan <sudvijayr(a)gmail.com> wrote: > We are doing prototype so much change have gone into kernel , we are > finding it difficult to upgrade to latest immediately > However I ran through the code once again, indeed kernel handles it > down_write(&mm->mmap_sem); in coredump_wait() makes sure the second > coredump is stopped and returns negative for core_waiters > > I will debug further, thanks for confirming that kernel handles this scenario > > > On Tue, Feb 17, 2015 at 8:42 PM, Greg KH <greg(a)kroah.com> wrote: >> On Tue, Feb 17, 2015 at 07:11:55PM +0530, Sudharsan Vijayaraghavan wrote: >>> Hi All, >>> >>> We are running 3.8 kernel. >> >> That's pretty old and obsolete, why are you stuck with that version? >> >>> I have a unique scenario, where we hit on several issues in do_coredump. >>> We have a SMP system with thousands of cores, one pthread is tied to >>> one core. The main process containing these pthreads runs in the first >>> core. >>> >>> Here is the issue # 1 >>> When one of threads core dump, we enter into do_coredump(), now one >>> other thread in same process running in a different >>> core can as well core dump(before SIGKILL was delivered to it as a >>> consequence of first core dump) >>> This gives way to entering into do_coredump more than once. >>> Once we have two guys entering do_coredump() one can kill other with SIGKILL >>> the result is completely unpredictable. No guarantee we will have two >>> core files generated in the end >>> >>> Linux kernel does not seem to handle it at all. >>> Adding a spin lock within do_coredump() will solve the case of >>> multiple entries into do_coredump() >>> >>> I want to know whether Linux kernel really does not handle the above >>> case or am I missing something? >> >> Odd, we should handle this just fine, try emailing the developers >> responsible for this code and cc: the linux-kernel mailing list so they >> can work it out. >> >> thanks, >> >> greg k-h

2 2

unlikely compiler flag propagation
by Matthias Brugger Feb. 19, 2015

Feb. 19, 2015

Hi all, I have a question about the unlikely compiler flag. When a called function is only returns an error with the unlikely flag set, should I set the unlikely compiler flag for the return value check in the callee as well? For example: int function_one(int *list, int num_elements) { int i; for (i =0; i < num_elements; i++) { if (unlikely(check_element(list + i))) return 1; } [...] return 0; } int function_two(...) { [...] if (function_one(list, num)) return -1; } So my question is, if function_two should instead implement: if (unlikely(function_one(list, num)) Or does the unlikely compiler flag propagate to calling functions? Thanks a lot, Matthias -- motzblog.wordpress.com

5 7

executing insmod hangs the entire os
by noyb noybee Feb. 18, 2015

Feb. 18, 2015

When I execute insmod for the kernel module object file of the following C code, the entire system hangs. The module replaces the reference to original chroot system call with a new one in the sys_call_table. The syscall_table address is correct as per System.map(which returns 2 values for sys_call_table, surprisingly). I am on a VM running CentOS 6.6 with kernel version 2.6.32-504. #include <linux/init.h> #include <linux/module.h> #include <linux/kernel.h> #include <linux/errno.h> #include <linux/types.h> #include <linux/unistd.h> #include <asm/cacheflush.h> #include <asm/page.h> #include <asm/current.h> #include <linux/sched.h> #include <linux/kallsyms.h> unsigned long *syscall_table = (unsigned long *)0xffffffff81600560; asmlinkage int (*original_chroot)(const char __user *); asmlinkage int new_chroot(const char __user *filename){ printk(KERN_ALERT "CHROOT HIJACKED"); return (*original_chroot)(filename); } static int init(void) { printk(KERN_ALERT "\nHIJACK INIT\n"); original_chroot = (void *)syscall_table[__NR_chroot]; syscall_table[__NR_chroot] = new_chroot; return 0; } static void exit(void) { syscall_table[__NR_chroot] = original_chroot; printk(KERN_ALERT "MODULE EXIT\n"); } module_init(init); module_exit(exit);

3 6

Mail client for android
by Ricardo Ribalda Delgado Feb. 17, 2015

Feb. 17, 2015

Hello: Since my first contribution I have used gmail as my mail provider. For sending patches I happily use git send-email and when I interact with a kverg list I use my gmail account via web (making sure that the mail is sent in text mode). But when I am travelling and I need to reply to a list, the gmail client decides on its own if it should use html or text reply. What client do you use in android, that can use a gmail account and play nice with the lists? Thanks! -- Ricardo Ribalda

2 1

Linux do_coredump() and SMP systems
by Sudharsan Vijayaraghavan Feb. 17, 2015

Feb. 17, 2015

Hi All, We are running 3.8 kernel. I have a unique scenario, where we hit on several issues in do_coredump. We have a SMP system with thousands of cores, one pthread is tied to one core. The main process containing these pthreads runs in the first core. Here is the issue # 1 When one of threads core dump, we enter into do_coredump(), now one other thread in same process running in a different core can as well core dump(before SIGKILL was delivered to it as a consequence of first core dump) This gives way to entering into do_coredump more than once. Once we have two guys entering do_coredump() one can kill other with SIGKILL the result is completely unpredictable. No guarantee we will have two core files generated in the end Linux kernel does not seem to handle it at all. Adding a spin lock within do_coredump() will solve the case of multiple entries into do_coredump() I want to know whether Linux kernel really does not handle the above case or am I missing something? Please clarify Issue # 2: Within do_coredump() SIGKILL is sent to all threads in process other than the one running core dump. There is no guarantee that SIGKILL will be immediately received by all threads in the process, which means the state of threads (particularly backtrace per thread) can be lot of different now when compared to the time at which offending thread initiated a coredump. This is in turn means the core dump generated, will have a backtrace per thread, which is not accurate Please confirm my understanding, advice on how this problem can be solved Thanks, Sudharsan

2 1

What is the kernel log buffer format
by manty kuma Feb. 17, 2015

Feb. 17, 2015

Hi all, What is the format of the kernel log buffer? I am expecting something like this: <<HEADER>> <<DATA>> <<END> I am looking at an issue where kernel logs are getting corrupted. Knowing the structure will help me. Also, any other tips would be very helpful. Regards, Manty

1 0

error: missing syscalls when compiling kernel module
by noyb noybee Feb. 17, 2015

Feb. 17, 2015

Hi all. This is the error that I am getting when I try to compile a loadable module(it tries to intercept a system call by directly editing the sys_call_table in the memory):- make -C /lib/modules/2.6.32-504.1.3.el6.x86_64/build M= modules make[1]: Entering directory `/usr/src/kernels/2.6.32-504.1.3.el6.x86_64' make[2]: Nothing to be done for `all'. CHK include/linux/version.h CHK include/linux/utsrelease.h SYMLINK include/asm -> include/asm-x86 make[2]: *** No rule to make target `missing-syscalls'. Stop. make[1]: *** [prepare0] Error 2 make[1]: Leaving directory `/usr/src/kernels/2.6.32-504.1.3.el6.x86_64' make: *** [all] Error 2 The Makefile is:- obj-m +:= hijack.o PWD := $(pwd) all: make -C /lib/modules/2.6.32-504.1.3.el6.x86_64/build M=$(PWD) modules clean: make -C /lib/modules/2.6.32-504.1.3.el6.x86_64/build M=$(PWD) clean I have googled the error but most links are unresolved. I am trying to install this against the currently running kernel and not the complete source.

5 4

Maintainer of .mailmap and CREDITS
by Ricardo Ribalda Delgado Feb. 17, 2015

Feb. 17, 2015

Hello I did send this two patches: [PATCH] CREDITS: Add Ricardo Ribalda Delgado https://lkml.org/lkml/2015/2/9/296 and [PATCH] .mailmap: Add Ricardo Ribalda https://lkml.org/lkml/2015/2/9/295 to the output of get_maintainers.pl Unfortunately, nobody has payed much attention :). Is there anybody in charge of this files, or a malining list to ping? Are these patches candidates for trivial patches? Thanks! -- Ricardo Ribalda

3 2