e0969e0c9b
The implementation of system call tracing in the kernel has a couple of ordering problems: - the validity of the system call number is checked before calling out to system call tracing code, and should be done after - the system call number used when tracing is the one the system call was invoked with, while the system call tracing code can legitimatly change the call number (for example strace permutes fork into clone) This patch fixes both of these problems, and also reoders the code slightly to make the direct path through the code the common case. Signed-off-by: Stuart Menefy <stuart.menefy@st.com> Signed-off-by: Paul Mundt <lethal@linux-sh.org> |
||
---|---|---|
.. | ||
cpu | ||
timers | ||
vsyscall | ||
apm.c | ||
asm-offsets.c | ||
cf-enabler.c | ||
cpufreq.c | ||
early_printk.c | ||
entry-common.S | ||
head.S | ||
init_task.c | ||
io_generic.c | ||
io.c | ||
irq.c | ||
kgdb_jmp.S | ||
kgdb_stub.c | ||
machine_kexec.c | ||
Makefile | ||
module.c | ||
pm.c | ||
process.c | ||
ptrace.c | ||
relocate_kernel.S | ||
semaphore.c | ||
setup.c | ||
sh_bios.c | ||
sh_ksyms.c | ||
signal.c | ||
smp.c | ||
sys_sh.c | ||
syscalls.S | ||
time.c | ||
traps.c | ||
vmlinux.lds.S |