- kernelnewbies@kernelnewbies.org - Kernelnewbies Mailing Lists

Help with learning scheduler O(1)
by Meyer Lansky March 13, 2015

March 13, 2015

Hello all continue to study the scheduler is O(1 ) there was one difficulty with the string 2 on 3, in that these blocking occurs after lines ? a reference to the code that study - http://lxr.free-electrons.com/source/kernel/sched.c?v=2.6.36 what I learn on line 4682 to 4695. 1 rcu_read_lock(); 2 pcred = __task_cred(p); 3 match = (cred->euid == pcred->euid || cred->euid == pcread->uid); 4 rcu_read_unlock(); I ask because in the comments there is not described . Thank.

2 1

Missing sys_enter_open events
by sahil aggarwal March 13, 2015

March 13, 2015

Hi all I am doing ftrace to track files being opened and mapped by process but sys_enter_open is getting missed and getname is being traced which is called by sys_open only. What is happening.? mem-2474 [005] 681774.985957: getnameprobe: (sys_execve+0x21/0x5a <- getname) arg1="bin/mem" mem-2474 [001] 683183.894867: sync_read: (vfs_read+0xab/0x107 <- do_sync_read) arg1=80 mem-2474 [001] 683183.894900: sync_read: (vfs_read+0xab/0x107 <- do_sync_read) arg1=1c0 mem-2474 [001] 683183.894902: sync_read: (vfs_read+0xab/0x107 <- do_sync_read) arg1=1c mem-2474 [001] 683183.894913: sync_read: (vfs_read+0xab/0x107 <- do_sync_read) arg1=80 mem-2474 [001] 683183.895113: sync_read: (vfs_read+0xab/0x107 <- do_sync_read) arg1=188 mem-2474 [001] 683183.895275: getnameprobe: (do_sys_open+0x3b/0x105 <- getname) arg1="/etc/ld.so.cache" mem-2474 [001] 683183.895316: getnameprobe: (do_sys_open+0x3b/0x105 <- getname) arg1="/lib/libc.so.6" mem-2474 [001] 683183.895321: sys_read(fd: 4, buf: 7fff8461a0f8, count: 340) mem-2474 [001] 683183.895328: sync_read: (vfs_read+0xab/0x107 <- do_sync_read) arg1=340

1 0

I am new here
by Leonardo Ariel Carreras Rodríguez March 12, 2015

March 12, 2015

Hello! I am new here, and I want to know how to get started, maybe with the simplest way available. Regards, Leonardo

6 8

Re: I am new here
by Jonathan Jin March 12, 2015

March 12, 2015

Leonardo Ariel Carreras Rodríguez <zc77leo(a)gmail.com> writes: > In particular, I am asking about how much time do you have to wait > until you get accepted to join the challenge. If s/he has to approve > each mail address, this will take a lot of time. Again, it varies. Challenge joins shouldn't take too long, as I imagine that processing those emails was one of the first jobs that got automated. If I remember correctly, my registration had a response time on the order of a couple of hours. Side note: don't top-post. It's very frowned upon in most mailing lists. See here: http://kernelnewbies.org/mailinglistguidelines. -- Jonathan Jin jjin.me github.com/jinnovation linkedin.com/in/jonathanjin

1 0

RCU is not safe enough to protect __d_lookup ?
by Rock Lee March 12, 2015

March 12, 2015

Hi, all: I am studying the dcache in VFS recently . I've found that hlist of dcache is already protected by RCU in __d_lookup.Is it necessary for the function -- d_lookup using sequence lock to protect the hlist again? struct dentry * d_lookup(struct dentry * parent, struct qstr * name){ struct dentry * dentry = NULL; unsigned long seq; do { seq = read_seqbegin(&rename_lock); dentry = __d_lookup(parent, name); if (dentry) break; } while (read_seqretry(&rename_lock, seq)); return dentry;} Thanks in advance--------------------------------Rock Lee

1 0

RCU is not safe enough to protect __d_lookup ?
by Rock Lee March 12, 2015

March 12, 2015

Hi, all: I am studying the dcache in VFS recently . I've found that hlist of dcache is already protected by RCU in __d_lookup.Is it necessary for the function -- d_lookup using sequence lock to protect the hlist again? struct dentry * d_lookup(struct dentry * parent, struct qstr * name){ struct dentry * dentry = NULL; unsigned long seq; do { seq = read_seqbegin(&rename_lock); dentry = __d_lookup(parent, name); if (dentry) break; } while (read_seqretry(&rename_lock, seq)); return dentry;} Thanks in advance--------------------------------Rock Lee

1 0

Tracing allocators of virtual memory and main memory
by sahil aggarwal March 12, 2015

March 12, 2015

Hi all If i want to trace the allocators of virtual memory and main memory of process using ftrace on which functions i need to enable the probe.? Currently i have enabled sys_mmap,sys_brk for virtual allocations and mm_page_alloc, kmalloc, kmem_cache_alloc for main memory allocation. Will this give me whole picture of virtual and main memory allocations.? Thanks Regards Sahil

4 5

Kernel Locking Question
by Nicholas Krause March 12, 2015

March 12, 2015

Greetings All, After learning about kernel locking and how atomic operations are good in terms of improving CPU scalability in terms of making a critical region much smaller. I was wondering of how this would improve file system scalability and reliability if implemented in file system code for btrfs worker threads. Nick -- Sent from my Android device with K-9 Mail. Please excuse my brevity.

3 2

Kernel preemption
by Rock Lee March 12, 2015

March 12, 2015

Hi, guys: I have a device with only one CPU. I am wondering whether there will be only one task running on CPU and this task will not be interrupted by scheduler if I make the kernel non-preemptive ? At this situation, whether the time splice will not exist? Regards --- Rock Lee

1 0

confusing code....whats the point of this construct ?
by Nicholas Mc Guire March 11, 2015

March 11, 2015

HI ! Trying to understand the intent of this code construct drivers/net/wireless/ath/ath10k/mac.c:ath10k_flush() <snip> ret = wait_event_timeout(ar->htt.empty_tx_wq, ({ bool empty; spin_lock_bh(&ar->htt.tx_lock); empty = (ar->htt.num_pending_tx == 0); spin_unlock_bh(&ar->htt.tx_lock); skip = (ar->state == ATH10K_STATE_WEDGED) || test_bit(ATH10K_FLAG_CRASH_FLUSH, &ar->dev_flags); (empty || skip); }), ATH10K_FLUSH_TIMEOUT_HZ); <snip> FYI include/linux/wait.h:wait_event_timeout() <snip> * wait_event_timeout - sleep until a condition gets true or a timeout elapses * @wq: the waitqueue to wait on * @condition: a C expression for the event to wait for * @timeout: timeout, in jiffies <snip> So the wait_event_timeout condition here ends up being (empty || skip) but what is the point of puting this code into the parameter list of wait_event_timeout() ? Would it not be equivalent to: bool empty; ... spin_lock_bh(&ar->htt.tx_lock); empty = (ar->htt.num_pending_tx == 0); spin_unlock_bh(&ar->htt.tx_lock); skip = (ar->state == ATH10K_STATE_WEDGED) || test_bit(ATH10K_FLAG_CRASH_FLUSH, &ar->dev_flags); ret = wait_event_timeout(ar->htt.empty_tx_wq, (empty || skip), ATH10K_FLUSH_TIMEOUT_HZ); What am I missing here ? thx! hofrat

6 12