[<prev] [next>] [thread-next>] [day] [month] [year] [list]
Message-ID: <45776D4F.6070806@gmail.com>
Date: Thu, 07 Dec 2006 02:24:08 +0059
From: Jiri Slaby <jirislaby@...il.com>
To: Eric Fox <efox@...steinindustries.com>
CC: Linux Kernel Mailing List <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH 1/1] Char: isicom, fix card locking
Eric Fox wrote:
> Patch works if I disable SMP in kernel.
Aha, and it didn't work without the patch with SMP disabled?
> strace setserial -g /dev/ttyM0
> execve("/bin/setserial", ["setserial", "-g", "/dev/ttyM0"], [/* 17 vars
> */]) = 0
> uname({sys="Linux", node="dialin-0.vab.com", ...}) = 0
> brk(0) = 0x804d000
> access("/etc/ld.so.preload", R_OK) = -1 ENOENT (No such file or
> directory)
> open("/etc/ld.so.cache", O_RDONLY) = 3
> fstat64(3, {st_mode=S_IFREG|0644, st_size=19713, ...}) = 0
> old_mmap(NULL, 19713, PROT_READ, MAP_PRIVATE, 3, 0) = 0xb7f58000
> close(3) = 0
> open("/lib/tls/libc.so.6", O_RDONLY) = 3
> read(3, "\177ELF\1\1\1\0\0\0\0\0\0\0\0\0\3\0\3\0\1\0\0\0\320\36"..., 512) =
> 512
> fstat64(3, {st_mode=S_IFREG|0755, st_size=1454802, ...}) = 0
> old_mmap(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1,
> 0) = 0xb7f57000
> old_mmap(0x39d000, 1223900, PROT_READ|PROT_EXEC, MAP_PRIVATE|MAP_DENYWRITE,
> 3, 0) = 0x39d000
> old_mmap(0x4c2000, 16384, PROT_READ|PROT_WRITE,
> MAP_PRIVATE|MAP_FIXED|MAP_DENYWRITE, 3, 0x124000) = 0x4c2000
> old_mmap(0x4c6000, 7388, PROT_READ|PROT_WRITE,
> MAP_PRIVATE|MAP_FIXED|MAP_ANONYMOUS, -1, 0) = 0x4c6000
> close(3) = 0
> old_mmap(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1,
> 0) = 0xb7f56000
> mprotect(0x4c2000, 4096, PROT_READ) = 0
> mprotect(0x399000, 4096, PROT_READ) = 0
> set_thread_area({entry_number:-1 -> 6, base_addr:0xb7f566c0, limit:1048575,
> seg_32bit:1, contents:0, read_exec_only:0, limit_in_pages:1,
> seg_not_present:0, 0munmap(0xb7f58000, 19713) = 0
> open("/dev/ttyM0", O_RDWR|O_NONBLOCK) = 3
> ioctl(3, TIOCGSERIAL, 0xbf81c280) = 0
In one of these two there should lie the problem in SMP mode. The latter is OK
(no locking), but the former -- hmm, it gets stuck: I guess, we have a recursive
locking example here (I can see one of many in the driver now):
isicom_open()
{
...
isicom_setup_port()
...
}
isicom_setup_port()
{
...
spin_lock_irqsave(&card->card_lock, flags);
...
isicom_config_port();
...
spin_unlock_irqrestore(&card->card_lock, flags);
}
isicom_config_port()
{
...
lock_card()
...
unlock_card()
...
}
lock_card()
{
...
spin_lock_irqsave(&card->card_lock, card->flags);
...
}
unlock_card()
{
spin_unlock_irqrestore(&card->card_lock, card->flags);
}
Going to fix these.
thanks,
--
http://www.fi.muni.cz/~xslaby/ Jiri Slaby
faculty of informatics, masaryk university, brno, cz
e-mail: jirislaby gmail com, gpg pubkey fingerprint:
B674 9967 0407 CE62 ACC8 22A0 32CC 55C3 39D4 7A7E
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Powered by blists - more mailing lists