May 2013 - kernelnewbies@kernelnewbies.org - Kernelnewbies Mailing Lists

Emulating cpu page protection in userspace
by Manish Katiyar May 16, 2013

May 16, 2013

Hi, This is really not a linux kernel question, but I guess there are enough OS specialists on this thread that someone can point me in the right direction. Some context ---------------------- On normal hardware machines its possible to change the page table flags to make a particular 4K page area as readonly or readwrite. This is generally used for protecting against memory corruptions. Now even if you have READONLY flags set in the page table, its possible to change the control register of a particular cpu (in a multicore system) such that, that cpu will not cause a fault/trap if you try to modify the contents of the page. What this means is let's say I have 4 cpus, and I change the control register on cpu-1, only cpu-1 will see the page as readwrite and others will see it as readonly. I've seen that this behavior is honoured in virtualised vmware environments too, so I'm assuming that this cpu behavior is getting simulated somehow. My problem is that I want to simulate the same behavior in userspace, where my 4 cpus map to 4 pthreads. I change the permissions of my memory area using mprotect() but then I also want to have a similar capability where I can flip/change something in one of the threads and that thread should be able to modify the region, while for other pthreads it is still a readonly page. Any suggestions how vmware or other virtualised environments do this, or is this even possible ? Any pointers to the code is appreciated. (Sorry no locking/synchronisation solutions pls). -- Thanks - Manish

4 3

how to debug thread stuck in filemap_fault
by shuang liu May 16, 2013

May 16, 2013

Hi, there, some user space threads, say thread-X stuck in sleep_on_page_killable @ filemap_fault(). my question is : in normal flow, how to wakeup thread-X from wait_on_page_locked_killable() ? How can I go further to debug this issue? I try to use ftrace function_graph to trace the flow, but it can only log EXPORT_SYMBOL_GPL() functions. I am struggling with huge memory and file system source code... hope anyone can give me some practical tips. Call Trace: [<c1895675>] schedule+0x35/0x50 [<c1895708>] io_schedule+0x78/0xb0 [<c12c95fd>] sleep_on_page_killable+0xd/0x40 [<c1895f3f>] __wait_on_bit+0x4f/0x70 [<c12cbcae>] wait_on_page_bit_killable+0x8e/0x90 [<c12cbd42>] __lock_page_or_retry+0x92/0xb0 [<c12cc041>] filemap_fault+0x2e1/0x400 [<c12e55be>] __do_fault+0x3e/0x390 [<c12e72ac>] handle_pte_fault+0x7c/0x930 [<c12e7bf3>] handle_mm_fault+0x93/0xc0 [<c189a98e>] do_page_fault+0x10e/0x4b0 [<c189881b>] error_code+0x5f/0x64 here is the kernel code in filemap.c static inline int wait_on_page_locked_killable(struct page *page) { if (PageLocked(page)) return wait_on_page_bit_killable(page, PG_locked); return 0; } here is ftrace result: 3) | filemap_fault() { 3) | find_get_page() { 3) 0.259 us | __rcu_read_lock(); 3) 0.217 us | __rcu_read_unlock(); 3) 7.432 us | } 3) 0.244 us | __might_sleep(); 3) + 12.206 us | }

2 1

Re: how do i read a block
by Shampavman May 15, 2013

May 15, 2013

Sorry imeant 3blocks Sankar P <sankar.curiosity(a)gmail.com> wrote: >On Mon, May 13, 2013 at 1:35 PM, shampavman <shampavman.cg(a)gmail.com> wrote: >> Hi all, >> >> Supposing i create a file of size 10K, it will occupy 2 blocks (4K each). >> Now if i want to read only 1 block from it how can i do it? >> > >It will require 3 blocks. > > >> read(fd, buf, 4096) ; >> would this mean i would read the first block and all its contents? >> > >It may be read even further (readahead) > >If you are in kernel mode, and if you want to implement your own >filesystem and want to read a single block, you can use the sb_bread >function. > > >-- >Sankar P >http://psankar.blogspot.com

1 0

how do i read a block
by shampavman May 15, 2013

May 15, 2013

Hi all, Supposing i create a file of size 10K, it will occupy 2 blocks (4K each). Now if i want to read only 1 block from it how can i do it? read(fd, buf, 4096) ; would this mean i would read the first block and all its contents? Thanks

6 7

how to debug oom killer errors
by Prashant Shah May 15, 2013

May 15, 2013

Hi, How to decode the stats thrown by the oom killer ? Is there any good in depth documentation out there for this ? Node 0 DMA per-cpu: cpu 0 hot: high 0, batch 1 used:0 cpu 0 cold: high 0, batch 1 used:0 Node 0 DMA32 per-cpu: cpu 0 hot: high 186, batch 31 used:25 cpu 0 cold: high 62, batch 15 used:49 Node 0 Normal per-cpu: empty Node 0 HighMem per-cpu: empty Free pages: 3532kB (0kB HighMem) Active:4471 inactive:4099 dirty:1 writeback:0 unstable:0 free:883 slab:14568 mapped-file:542 mapped-anon:8049 pagetables:1033 Node 0 DMA free:368kB min:44kB low:52kB high:64kB active:0kB inactive:0kB present:9700kB pages_scanned:0 all_unreclaimable? yes lowmem_reserve[]: 0 632 632 632 Node 0 DMA32 free:3164kB min:3192kB low:3988kB high:4788kB active:17884kB inactive:16396kB present:647348kB pages_scanned:920345 all_unreclaims lowmem_reserve[]: 0 0 0 0 Node 0 Normal free:0kB min:0kB low:0kB high:0kB active:0kB inactive:0kB present:0kB pages_scanned:0 all_unreclaimable? no lowmem_reserve[]: 0 0 0 0 Node 0 HighMem free:0kB min:128kB low:128kB high:128kB active:0kB inactive:0kB present:0kB pages_scanned:0 all_unreclaimable? no lowmem_reserve[]: 0 0 0 0 Node 0 DMA: 30*4kB 7*8kB 2*16kB 1*32kB 0*64kB 1*128kB 0*256kB 0*512kB 0*1024kB 0*2048kB 0*4096kB = 368kB Node 0 DMA32: 81*4kB 53*8kB 65*16kB 23*32kB 0*64kB 1*128kB 0*256kB 1*512kB 0*1024kB 0*2048kB 0*4096kB = 3164kB Node 0 Normal: empty Node 0 HighMem: empty 8602 pagecache pages Swap cache: add 21037, delete 12988, find 320/473, race 0+1 Free swap = 940120kB Total swap = 1020116kB Out of memory: Killed process 2024, UID 51, (sendmail). Regards.

1 1

Re: user space device drivers
by Gergely Buday May 14, 2013

May 14, 2013

>> Are there serious user space drivers in Linux? Could you name a few? > > Printer drivers, scanner drivers, file system drivers etc. How can I find them? Are they in the kernel source tree? - Gergely

3 2

Re: user space device drivers
by jeshkumar555＠gmail.com May 14, 2013

May 14, 2013

Hi :), One more mechanism I know is passing real time signal to the userspace process using process PID. So whenever there is an interrupt in kernel module, it passes the signal to userspace process. You can handle the signal in US process. And I passed the PID to kernel module using IOCTL. Let me know if anyone finds problem in handling interrupts like this. Thanks Regards, Jeshwanth Bengaluru ----- Reply message ----- From: "Prabhu nath" <gprabhunath(a)gmail.com> Date: Tue, May 14, 2013 3:29 pm Subject: user space device drivers To: "Gergely Buday" <gbuday(a)gmail.com> Cc: "kernelnewbies" <kernelnewbies(a)kernelnewbies.org> One that I know is through proc interface where the interrupt info is lodged in a file in /proc and the user code keeps polling on this file. Exact use of this is to be looked for. On Tue, May 14, 2013 at 3:24 PM, Gergely Buday <gbuday(a)gmail.com> wrote: > Prabhu nath wrote: > > > But if the device has to generate an interrupt on the reception of the > data > > then it is best for the driver code to be in the kernel space waiting for > > the data, rather than in the user space because there is no efficient > > mechanism till now for the control to be transferred to the waiting user > > space code on the reception of the interrupt. > > What are the curently available mechanisms for an interrupt to > propagate into user space? > > Is there one that affects the scheduler? > > - Gergely > -- Regards, Prabhunath G Linux Trainer Bangalore

1 0

Re: user space device drivers
by Gergely Buday May 14, 2013

May 14, 2013

Prabhu nath wrote: > But if the device has to generate an interrupt on the reception of the data > then it is best for the driver code to be in the kernel space waiting for > the data, rather than in the user space because there is no efficient > mechanism till now for the control to be transferred to the waiting user > space code on the reception of the interrupt. What are the curently available mechanisms for an interrupt to propagate into user space? Is there one that affects the scheduler? - Gergely

3 2

Wake on Lan and pci_enable_wake()
by Andy Johnson May 14, 2013

May 14, 2013

Hello, As far as I know, Wake On Lan is disabled by default in Linux. Now I try this on p2p1: ethtool -i p2p1 ... driver: r8169 ... ethtool p2p1 .. Wake-on: g ... this means, according to the man, that Wake on Lan is enabled for MagicPacket packets on p2p1 Now I looked in the code of the driver and I cannot see anywhere a call to pci_enable_wake(); see drivers/net/ethernet/realtek/r8169.c So any ideas how come that the Wake on Lan is enabled for this driver after boot ? rgs, Andy

1 0

PCIe MSI smp_affinity for ARM architecture
by Amit Mahadik May 14, 2013

May 14, 2013

Hello, I have an ARM platform board. I need to change the smp_affinity of PCIe MSI interrupt. The drivers/pci/msi.c file does not any function for changing the smp_affinity. Any pointers as to how to change the affinity would be very helpful. Regards, Amit.

1 1