Re: [PATCH] powerpc/powermac: Fix low_sleep_handler with KUAP and KUEP
Christophe Leroy writes: > Le 11/09/2020 à 01:56, Michael Ellerman a écrit : >> Christophe Leroy writes: >>> low_sleep_handler() has an hardcoded restore of segment registers >>> that doesn't take KUAP and KUEP into account. >>> >>> Use head_32's load_segment_registers() routine instead. >>> >>> Signed-off-by: Christophe Leroy >>> Fixes: a68c31fc01ef ("powerpc/32s: Implement Kernel Userspace Access >>> Protection") >>> Fixes: 31ed2b13c48d ("powerpc/32s: Implement Kernel Userspace Execution >>> Prevention.") >>> Cc: sta...@vger.kernel.org >>> --- >>> arch/powerpc/platforms/powermac/sleep.S | 9 + >>> 1 file changed, 1 insertion(+), 8 deletions(-) >> >> Doesn't build? pmac32_defconfig, gcc 9.3.0: >> >> ld: arch/powerpc/platforms/powermac/sleep.o: in function `core99_wake_up': >> (.text+0x25c): undefined reference to `load_segment_registers' >> >> Missing _GLOBAL() presumably? > > Oops .. :( > > v2 sent out. Thanks. cheers
Re: [PATCH] powerpc/powermac: Fix low_sleep_handler with KUAP and KUEP
Le 11/09/2020 à 01:56, Michael Ellerman a écrit : Christophe Leroy writes: low_sleep_handler() has an hardcoded restore of segment registers that doesn't take KUAP and KUEP into account. Use head_32's load_segment_registers() routine instead. Signed-off-by: Christophe Leroy Fixes: a68c31fc01ef ("powerpc/32s: Implement Kernel Userspace Access Protection") Fixes: 31ed2b13c48d ("powerpc/32s: Implement Kernel Userspace Execution Prevention.") Cc: sta...@vger.kernel.org --- arch/powerpc/platforms/powermac/sleep.S | 9 + 1 file changed, 1 insertion(+), 8 deletions(-) Doesn't build? pmac32_defconfig, gcc 9.3.0: ld: arch/powerpc/platforms/powermac/sleep.o: in function `core99_wake_up': (.text+0x25c): undefined reference to `load_segment_registers' Missing _GLOBAL() presumably? Oops .. :( v2 sent out. Thanks Christophe
Re: [PATCH] powerpc/powermac: Fix low_sleep_handler with KUAP and KUEP
Christophe Leroy writes: > low_sleep_handler() has an hardcoded restore of segment registers > that doesn't take KUAP and KUEP into account. > > Use head_32's load_segment_registers() routine instead. > > Signed-off-by: Christophe Leroy > Fixes: a68c31fc01ef ("powerpc/32s: Implement Kernel Userspace Access > Protection") > Fixes: 31ed2b13c48d ("powerpc/32s: Implement Kernel Userspace Execution > Prevention.") > Cc: sta...@vger.kernel.org > --- > arch/powerpc/platforms/powermac/sleep.S | 9 + > 1 file changed, 1 insertion(+), 8 deletions(-) Doesn't build? pmac32_defconfig, gcc 9.3.0: ld: arch/powerpc/platforms/powermac/sleep.o: in function `core99_wake_up': (.text+0x25c): undefined reference to `load_segment_registers' Missing _GLOBAL() presumably? cheers > diff --git a/arch/powerpc/platforms/powermac/sleep.S > b/arch/powerpc/platforms/powermac/sleep.S > index f9a680fdd9c4..51bfdfe85058 100644 > --- a/arch/powerpc/platforms/powermac/sleep.S > +++ b/arch/powerpc/platforms/powermac/sleep.S > @@ -294,14 +294,7 @@ grackle_wake_up: >* we do any r1 memory access as we are not sure they >* are in a sane state above the first 256Mb region >*/ > - li r0,16 /* load up segment register values */ > - mtctr r0 /* for context 0 */ > - lis r3,0x2000 /* Ku = 1, VSID = 0 */ > - li r4,0 > -3: mtsrin r3,r4 > - addir3,r3,0x111 /* increment VSID */ > - addis r4,r4,0x1000/* address of next segment */ > - bdnz3b > + bl load_segment_registers > sync > isync > > -- > 2.25.0
[PATCH] powerpc/powermac: Fix low_sleep_handler with KUAP and KUEP
low_sleep_handler() has an hardcoded restore of segment registers that doesn't take KUAP and KUEP into account. Use head_32's load_segment_registers() routine instead. Signed-off-by: Christophe Leroy Fixes: a68c31fc01ef ("powerpc/32s: Implement Kernel Userspace Access Protection") Fixes: 31ed2b13c48d ("powerpc/32s: Implement Kernel Userspace Execution Prevention.") Cc: sta...@vger.kernel.org --- arch/powerpc/platforms/powermac/sleep.S | 9 + 1 file changed, 1 insertion(+), 8 deletions(-) diff --git a/arch/powerpc/platforms/powermac/sleep.S b/arch/powerpc/platforms/powermac/sleep.S index f9a680fdd9c4..51bfdfe85058 100644 --- a/arch/powerpc/platforms/powermac/sleep.S +++ b/arch/powerpc/platforms/powermac/sleep.S @@ -294,14 +294,7 @@ grackle_wake_up: * we do any r1 memory access as we are not sure they * are in a sane state above the first 256Mb region */ - li r0,16 /* load up segment register values */ - mtctr r0 /* for context 0 */ - lis r3,0x2000 /* Ku = 1, VSID = 0 */ - li r4,0 -3: mtsrin r3,r4 - addir3,r3,0x111 /* increment VSID */ - addis r4,r4,0x1000/* address of next segment */ - bdnz3b + bl load_segment_registers sync isync -- 2.25.0