On Tue, Jan 17, 2017 at 12:43 PM, Segher Boessenkool <seg...@kernel.crashing.org> wrote: > Currently, on PowerPC, code compiled with -fstack-protector will load > the canary from -0x7010(13) (for -m64) or from -0x7008(2) (for -m32) > if GCC was compiled against GNU libc 2.4 or newer or some other libc > that supports -fstack-protector, and from the global variable > __stack_chk_guard otherwise. > > This does not work well for Linux and other OS kernels and similar. > For such non-standard applications, this patch creates a few new > command-line options. The relevant new use cases are: > > -mstack-protector-guard=global > Use the __stack_chk_guard variable, no matter how this GCC was > configured. > > -mstack-protector-guard=tls > Use the canary from TLS. This will error out if this GCC was built > with a C library that does not support it. > > -mstack-protector-guard=tls -mstack-protector-register=<reg> > -mstack-protector-offset=<offset> > Load the canary from offset <off> from base register <reg>. > > > Bootstrap and test running. Is this okay for trunk? > > > Segher > > > 2017-01-17 Segher Boessenkool <seg...@kernel.crashing.org> > > PR target/78875 > > * config/rs6000/rs6000-opts.h (stack_protector_guard): New enum. > * config/rs6000/rs6000.c (rs6000_option_override_internal): Handle > the new options. > * config/rs6000/rs6000.md (stack_protect_set): Handle the new more > flexible settings. > (stack_protect_test): Ditto. > * config/rs6000/rs6000.opt (mstack-protector-guard=, > mstack-protector-guard-reg=, mstack-protector-guard-offset=): New > options. > * doc/invoke.texi (Option Summary) [RS/6000 and PowerPC Options]: > Add -mstack-protector-guard=, -mstack-protector-guard-reg=, and > -mstack-protector-guard-offset=. > (RS/6000 and PowerPC Options): Ditto. > > gcc/testsuite/ > * gcc.target/powerpc/ssp-1.c: New testcase. > * gcc.target/powerpc/ssp-2.c: New testcase.
Okay. Thanks, David
