strace of sctl-0.2.3 shows mmap DENIED
I think this is because I'm using a hardened kernel.
However, I'm seeing no mention of these mmap DENIED
within syslog. (I also have yet to debug this by using
another standard kernel.)
From what I'm seeing though, I can read the output from
the port as when the scanner activates on a frequency,
I will see a stream of "read" output from the active
strace of sctl.
In breif, looks like my kernel is denying memory writes
to the tty device.
(So, this data is just posted for other users to see
incase they run into a similar problem -- if I resolve
it by using my initial suspicions!)
$ strace sctl light
... begin clipping ...
fstat64(3, {st_mode=S_IFREG|0755, st_size=1218960,
...}) = 0
mmap2(NULL, 1154348, PROT_READ|PROT_EXEC,
MAP_PRIVATE|MAP_DENYWRITE, 3, 0) = 0x401d3000
mprotect(0x402e6000, 27948, PROT_NONE) = 0
mmap2(0x402e7000, 16384, PROT_READ|PROT_WRITE,
MAP_PRIVATE|MAP_FIXED|MAP_DENYWRITE, 3, 0x113) = 0x402e7000
mmap2(0x402eb000, 7468, PROT_READ|PROT_WRITE,
MAP_PRIVATE|MAP_FIXED|MAP_ANONYMOUS, -1, 0) = 0x402eb000
close(3) = 0
mmap2(NULL, 4096, PROT_READ|PROT_WRITE,
MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0xb7e1d000
mprotect(0x402e7000, 4096, PROT_READ) = 0
set_thread_area({entry_number:-1 -> 6,
base_addr:0xb7e1d8e0, limit:1048575, seg_32bit:1,
contents:0, read_exec_only:0, limit_in_pages:1,
seg_not_present:0, useable:1}) = 0
... end clipping ...
Logged In: YES
user_id=348797
Seems I'm seeing a hang within command.c at "get "raw" mode"
I'm seeing MD being written to the device and hanging while
I am seeing output when scanner reacts to an active channel
(ie. "status = get_mode(fd);" ?)
strace:
write(4, "MD\r", 3) = 3
read(4, [hangs and continues when scanner is activated with
output]
Logged In: YES
user_id=348797
Ok. I got sctl working.
I'm almost positve this is a grsec (hardened) kernel problem
because it isn't noting the mmap violation within syslog and
it can only be seen with strace.
Logged In: YES
user_id=348797
Ok. Verified. It is with the grsec hardened kernel that I
run into this problem.
I believe this is verified by the error to write protected
memory by grsec modules which *should* dump errors to syslog
telling users that such a violation occured (not a sctl
problem).
Also, I have a sony ericsson gc82 (pcmcia modem using
serial_cs.o module. It would appear with this card plugged
in, it takes resournces from my on board COM0 port. Nothing
unusual though. A simple Belkin tty>usb adapter will
probably resolve the resource conflict. I basically just
get a hang with writing the status command to teh scanner...
no other unusual errors. I'm wondering if there's a syscall
that could be implemented to advise users of the potential
resource conflict? (no biggy anyways, think most people will
just get a serial>usb adapter)