Re: Linux kernel Books
On Fri, Mar 23, 2018 at 10:00 AM, Milad Kahsariwrote: > This book illustrates virtual memory space in a reasonably clear way, easy > to follow with lots of diagrams, which I did not find in other books. > Although it states IA-64, the concept is pretty much the same if you are > looking for IA-32, excellent book, definitely recommend it! :) It's worth to note that IA-64 [1] is not x86-64 [2]. IA-64 is the Itanium ISA. Is there still anyone using this? :) Best regards, [1] https://en.wikipedia.org/wiki/IA-64 [2] https://en.wikipedia.org/wiki/X86-64 -- Augusto Mecking Caringi ___ Kernelnewbies mailing list Kernelnewbies@kernelnewbies.org https://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies
Re: Hang in FUTEX_WAKE_PRIVATE after clock_gettime
On Sat, 24 Mar 2018 04:43:42 -0400, Jeffrey Walton said: > My question is, is a futex with a 2147483647 or -1 expected? Where > should I be looking to begin to narrow the problem down? 'man 7 futex' says: Semantics Any futex operation starts in user space, but it may be necessary to communicate with the kernel using the futex(2) system call. To "up" a futex, execute the proper assembler instructions that will cause the host CPU to atomically increment the integer. Afterward, check if it has in fact changed from 0 to 1, in which case there were no waiters and the operation is done. This is the noncontended case which is fast and should be common. In the contended case, the atomic increment changed the counter from -1 (or some other negative number). If this is detected, there are wait- ers. User space should now set the counter to 1 and instruct the ker- nel to wake up any waiters using the FUTEX_WAKE operation. Waiting on a futex, to "down" it, is the reverse operation. Atomically decrement the counter and check if it changed to 0, in which case the operation is done and the futex was uncontended. In all other circum- stances, the process should set the counter to -1 and request that the kernel wait for another process to up the futex. This is done using the FUTEX_WAIT operation. The futex(2) system call can optionally be passed a timeout specifying how long the kernel should wait for the futex to be upped. In this case, semantics are more complex and the programmer is referred to futex(2) for more details. The same holds for asynchronous futex waiting. which pretty much says that a negative value means there's somebody waiting on the futex. pgpLKt41FqBMx.pgp Description: PGP signature ___ Kernelnewbies mailing list Kernelnewbies@kernelnewbies.org https://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies
Hang in FUTEX_WAKE_PRIVATE after clock_gettime
Hi Everyone, I'm working with Botan (https://botan.randombit.net/). It is a TLS security library written in C++11. On an old Pentium III machine the self tests are hanging. Other platforms are OK. The strace tail is shown below. I am not that familiar with this type of debugging, and I am looking for a direction to go. I've seen some similar problems on the web, including an indictment of clock_gettime, but the problems lack a good analysis of the problem. My question is, is a futex with a 2147483647 or -1 expected? Where should I be looking to begin to narrow the problem down? Thanks in advance. ** $ sudo strace ./botan-test block execve("./botan-test", ["./botan-test", "block"], [/* 14 vars */]) = 0 brk(NULL) = 0x9af9000 access("/etc/ld.so.nohwcap", F_OK) = -1 ENOENT (No such file or directory) readlink("/proc/self/exe", "/home/jwalton/botan/botan-test", 4096) = 30 mmap2(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0xb7f4e000 ... stat64("src/tests/data/block/noekeon.vec", {st_mode=S_IFREG|0664, st_size=121320, ...}) = 0 stat64("src/tests/data/block/aria.vec", {st_mode=S_IFREG|0664, st_size=2663, ...}) = 0 stat64("src/tests/data/block/des.vec", {st_mode=S_IFREG|0664, st_size=28722, ...}) = 0 getdents(3, /* 0 entries */, 32768) = 0 close(3)= 0 open("src/tests/data/block/aes.vec", O_RDONLY|O_LARGEFILE) = 3 read(3, "# Test vectors from NIST CAVP AE"..., 8191) = 8191 clock_gettime(CLOCK_REALTIME, {1521879372, 559396333}) = 0 futex(0xb78b6a34, FUTEX_WAKE_PRIVATE, 2147483647) = 0 * This old machine runs Peppermint, which is a Ubuntu derivative. $ lsb_release -a No LSB modules are available. Distributor ID: Peppermint Description:Peppermint 8 Eight Release:8 Codename: xenial $ uname -a Linux coppermine 4.13.0-37-generic #42~16.04.1-Ubuntu SMP Wed Mar 7 16:02:25 UTC 2018 i686 i686 i686 GNU/Linux ___ Kernelnewbies mailing list Kernelnewbies@kernelnewbies.org https://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies