2020-03-30 02:24:14

by Jason Yan

[permalink] [raw]
Subject: [PATCH v5 0/6] implement KASLR for powerpc/fsl_booke/64

This is a try to implement KASLR for Freescale BookE64 which is based on
my earlier implementation for Freescale BookE32:
https://patchwork.ozlabs.org/project/linuxppc-dev/list/?series=131718&state=*

The implementation for Freescale BookE64 is similar as BookE32. One
difference is that Freescale BookE64 set up a TLB mapping of 1G during
booting. Another difference is that ppc64 needs the kernel to be
64K-aligned. So we can randomize the kernel in this 1G mapping and make
it 64K-aligned. This can save some code to creat another TLB map at
early boot. The disadvantage is that we only have about 1G/64K = 16384
slots to put the kernel in.

KERNELBASE

64K |--> kernel <--|
| | |
+--+--+--+ +--+--+--+--+--+--+--+--+--+ +--+--+
| | | |....| | | | | | | | | |....| | |
+--+--+--+ +--+--+--+--+--+--+--+--+--+ +--+--+
| | 1G
|-----> offset <-----|

kernstart_virt_addr

I'm not sure if the slot numbers is enough or the design has any
defects. If you have some better ideas, I would be happy to hear that.

Thank you all.

v4->v5:
Fix "-Werror=maybe-uninitialized" compile error.
Fix typo "similar as" -> "similar to".
v3->v4:
Do not define __kaslr_offset as a fixed symbol. Reference __run_at_load and
__kaslr_offset by symbol instead of magic offsets.
Use IS_ENABLED(CONFIG_PPC32) instead of #ifdef CONFIG_PPC32.
Change kaslr-booke32 to kaslr-booke in index.rst
Switch some instructions to 64-bit.
v2->v3:
Fix build error when KASLR is disabled.
v1->v2:
Add __kaslr_offset for the secondary cpu boot up.

Jason Yan (6):
powerpc/fsl_booke/kaslr: refactor kaslr_legal_offset() and
kaslr_early_init()
powerpc/fsl_booke/64: introduce reloc_kernel_entry() helper
powerpc/fsl_booke/64: implement KASLR for fsl_booke64
powerpc/fsl_booke/64: do not clear the BSS for the second pass
powerpc/fsl_booke/64: clear the original kernel if randomized
powerpc/fsl_booke/kaslr: rename kaslr-booke32.rst to kaslr-booke.rst
and add 64bit part

Documentation/powerpc/index.rst | 2 +-
.../{kaslr-booke32.rst => kaslr-booke.rst} | 35 ++++++-
arch/powerpc/Kconfig | 2 +-
arch/powerpc/kernel/exceptions-64e.S | 23 +++++
arch/powerpc/kernel/head_64.S | 13 +++
arch/powerpc/kernel/setup_64.c | 3 +
arch/powerpc/mm/mmu_decl.h | 23 +++--
arch/powerpc/mm/nohash/kaslr_booke.c | 91 +++++++++++++------
8 files changed, 147 insertions(+), 45 deletions(-)
rename Documentation/powerpc/{kaslr-booke32.rst => kaslr-booke.rst} (59%)

--
2.17.2


2020-03-30 02:31:56

by Jason Yan

[permalink] [raw]
Subject: [PATCH v5 3/6] powerpc/fsl_booke/64: implement KASLR for fsl_booke64

The implementation for Freescale BookE64 is similar as BookE32. One
difference is that Freescale BookE64 set up a TLB mapping of 1G during
booting. Another difference is that ppc64 needs the kernel to be
64K-aligned. So we can randomize the kernel in this 1G mapping and make
it 64K-aligned. This can save some code to creat another TLB map at
early boot. The disadvantage is that we only have about 1G/64K = 16384
slots to put the kernel in.

To support secondary cpu boot up, a variable __kaslr_offset was added in
first_256B section. This can help secondary cpu get the kaslr offset
before the 1:1 mapping has been setup.

Signed-off-by: Jason Yan <[email protected]>
Cc: Scott Wood <[email protected]>
Cc: Diana Craciun <[email protected]>
Cc: Michael Ellerman <[email protected]>
Cc: Christophe Leroy <[email protected]>
Cc: Benjamin Herrenschmidt <[email protected]>
Cc: Paul Mackerras <[email protected]>
Cc: Nicholas Piggin <[email protected]>
Cc: Kees Cook <[email protected]>
---
arch/powerpc/Kconfig | 2 +-
arch/powerpc/kernel/exceptions-64e.S | 10 ++++
arch/powerpc/kernel/head_64.S | 6 +++
arch/powerpc/kernel/setup_64.c | 3 ++
arch/powerpc/mm/mmu_decl.h | 20 ++++----
arch/powerpc/mm/nohash/kaslr_booke.c | 75 +++++++++++++++++++---------
6 files changed, 83 insertions(+), 33 deletions(-)

diff --git a/arch/powerpc/Kconfig b/arch/powerpc/Kconfig
index 497b7d0b2d7e..0c76601fdd59 100644
--- a/arch/powerpc/Kconfig
+++ b/arch/powerpc/Kconfig
@@ -564,7 +564,7 @@ config RELOCATABLE

config RANDOMIZE_BASE
bool "Randomize the address of the kernel image"
- depends on (FSL_BOOKE && FLATMEM && PPC32)
+ depends on (PPC_FSL_BOOK3E && FLATMEM)
depends on RELOCATABLE
help
Randomizes the virtual address at which the kernel image is
diff --git a/arch/powerpc/kernel/exceptions-64e.S b/arch/powerpc/kernel/exceptions-64e.S
index 1b9b174bee86..260cf1f1e71c 100644
--- a/arch/powerpc/kernel/exceptions-64e.S
+++ b/arch/powerpc/kernel/exceptions-64e.S
@@ -1378,6 +1378,7 @@ skpinv: addi r6,r6,1 /* Increment */
1: mflr r6
addi r6,r6,(2f - 1b)
tovirt(r6,r6)
+ add r6,r6,r19
lis r7,MSR_KERNEL@h
ori r7,r7,MSR_KERNEL@l
mtspr SPRN_SRR0,r6
@@ -1400,6 +1401,7 @@ skpinv: addi r6,r6,1 /* Increment */

/* We translate LR and return */
tovirt(r8,r8)
+ add r8,r8,r19
mtlr r8
blr

@@ -1528,6 +1530,7 @@ a2_tlbinit_code_end:
*/
_GLOBAL(start_initialization_book3e)
mflr r28
+ li r19, 0

/* First, we need to setup some initial TLBs to map the kernel
* text, data and bss at PAGE_OFFSET. We don't have a real mode
@@ -1570,6 +1573,12 @@ _GLOBAL(book3e_secondary_core_init)
cmplwi r4,0
bne 2f

+ li r19, 0
+#ifdef CONFIG_RANDOMIZE_BASE
+ LOAD_REG_ADDR_PIC(r19, __kaslr_offset)
+ ld r19,0(r19)
+ rlwinm r19,r19,0,0,5
+#endif
/* Setup TLB for this core */
bl initial_tlb_book3e

@@ -1602,6 +1611,7 @@ _GLOBAL(book3e_secondary_core_init)
lis r3,PAGE_OFFSET@highest
sldi r3,r3,32
or r28,r28,r3
+ add r28,r28,r19
1: mtlr r28
blr

diff --git a/arch/powerpc/kernel/head_64.S b/arch/powerpc/kernel/head_64.S
index ad79fddb974d..454129a3c259 100644
--- a/arch/powerpc/kernel/head_64.S
+++ b/arch/powerpc/kernel/head_64.S
@@ -104,6 +104,12 @@ __secondary_hold_acknowledge:
.8byte 0x0

#ifdef CONFIG_RELOCATABLE
+#ifdef CONFIG_RANDOMIZE_BASE
+ .globl __kaslr_offset
+__kaslr_offset:
+ .8byte 0x0
+#endif
+
/* This flag is set to 1 by a loader if the kernel should run
* at the loaded address instead of the linked address. This
* is used by kexec-tools to keep the the kdump kernel in the
diff --git a/arch/powerpc/kernel/setup_64.c b/arch/powerpc/kernel/setup_64.c
index e05e6dd67ae6..836e202dfd5b 100644
--- a/arch/powerpc/kernel/setup_64.c
+++ b/arch/powerpc/kernel/setup_64.c
@@ -67,6 +67,7 @@
#include <asm/kup.h>
#include <asm/early_ioremap.h>

+#include <mm/mmu_decl.h>
#include "setup.h"

int spinning_secondaries;
@@ -300,6 +301,8 @@ void __init early_setup(unsigned long dt_ptr)
/* Enable early debugging if any specified (see udbg.h) */
udbg_early_init();

+ kaslr_early_init(__va(dt_ptr), 0);
+
udbg_printf(" -> %s(), dt_ptr: 0x%lx\n", __func__, dt_ptr);

/*
diff --git a/arch/powerpc/mm/mmu_decl.h b/arch/powerpc/mm/mmu_decl.h
index 605129b5ccdf..6efbd7fd88a4 100644
--- a/arch/powerpc/mm/mmu_decl.h
+++ b/arch/powerpc/mm/mmu_decl.h
@@ -139,22 +139,16 @@ extern unsigned long calc_cam_sz(unsigned long ram, unsigned long virt,
extern void adjust_total_lowmem(void);
extern int switch_to_as1(void);
extern void restore_to_as0(int esel, int offset, void *dt_ptr, int bootcpu);
+#endif
void create_kaslr_tlb_entry(int entry, unsigned long virt, phys_addr_t phys);
extern int is_second_reloc;
-#endif
+extern unsigned long __kaslr_offset;
+extern unsigned int __run_at_load;

void reloc_kernel_entry(void *fdt, long addr);
extern void loadcam_entry(unsigned int index);
extern void loadcam_multi(int first_idx, int num, int tmp_idx);

-#ifdef CONFIG_RANDOMIZE_BASE
-void kaslr_early_init(void *dt_ptr, phys_addr_t size);
-void kaslr_late_init(void);
-#else
-static inline void kaslr_early_init(void *dt_ptr, phys_addr_t size) {}
-static inline void kaslr_late_init(void) {}
-#endif
-
struct tlbcam {
u32 MAS0;
u32 MAS1;
@@ -164,6 +158,14 @@ struct tlbcam {
};
#endif

+#ifdef CONFIG_RANDOMIZE_BASE
+void kaslr_early_init(void *dt_ptr, phys_addr_t size);
+void kaslr_late_init(void);
+#else
+static inline void kaslr_early_init(void *dt_ptr, phys_addr_t size) {}
+static inline void kaslr_late_init(void) {}
+#endif
+
#if defined(CONFIG_PPC_BOOK3S_32) || defined(CONFIG_FSL_BOOKE) || defined(CONFIG_PPC_8xx)
/* 6xx have BATS */
/* FSL_BOOKE have TLBCAM */
diff --git a/arch/powerpc/mm/nohash/kaslr_booke.c b/arch/powerpc/mm/nohash/kaslr_booke.c
index 6ebff31fefcc..24ad34641869 100644
--- a/arch/powerpc/mm/nohash/kaslr_booke.c
+++ b/arch/powerpc/mm/nohash/kaslr_booke.c
@@ -228,10 +228,11 @@ static __init unsigned long get_usable_address(const void *fdt,
unsigned long start,
unsigned long offset)
{
+ unsigned long unit = IS_ENABLED(CONFIG_PPC32) ? SZ_16K : SZ_64K;
unsigned long pa;
unsigned long pa_end;

- for (pa = offset; (long)pa > (long)start; pa -= SZ_16K) {
+ for (pa = offset; (long)pa > (long)start; pa -= unit) {
pa_end = pa + regions.kernel_size;
if (overlaps_region(fdt, pa, pa_end))
continue;
@@ -268,24 +269,34 @@ static unsigned long __init kaslr_legal_offset(void *dt_ptr, unsigned long rando
unsigned long index;
unsigned long offset;

- /*
- * Decide which 64M we want to start
- * Only use the low 8 bits of the random seed
- */
- index = random & 0xFF;
- index %= regions.linear_sz / SZ_64M;
-
- /* Decide offset inside 64M */
- offset = random % (SZ_64M - regions.kernel_size);
- offset = round_down(offset, SZ_16K);
+ if (IS_ENABLED(CONFIG_PPC32)) {
+ /*
+ * Decide which 64M we want to start
+ * Only use the low 8 bits of the random seed
+ */
+ index = random & 0xFF;
+ index %= regions.linear_sz / SZ_64M;
+
+ /* Decide offset inside 64M */
+ offset = random % (SZ_64M - regions.kernel_size);
+ offset = round_down(offset, SZ_16K);
+
+ while ((long)index >= 0) {
+ offset = memstart_addr + index * SZ_64M + offset;
+ start = memstart_addr + index * SZ_64M;
+ koffset = get_usable_address(dt_ptr, start, offset);
+ if (koffset)
+ break;
+ index--;
+ }
+ } else {
+ /* Decide kernel offset inside 1G */
+ offset = random % (regions.linear_sz - regions.kernel_size);
+ offset = round_down(offset, SZ_64K);

- while ((long)index >= 0) {
- offset = memstart_addr + index * SZ_64M + offset;
- start = memstart_addr + index * SZ_64M;
+ start = memstart_addr;
+ offset = memstart_addr + offset;
koffset = get_usable_address(dt_ptr, start, offset);
- if (koffset)
- break;
- index--;
}

if (koffset != 0)
@@ -325,6 +336,7 @@ static unsigned long __init kaslr_choose_location(void *dt_ptr, phys_addr_t size
else
pr_warn("KASLR: No safe seed for randomizing the kernel base.\n");

+#ifdef CONFIG_PPC32
ram = min_t(phys_addr_t, __max_low_memory, size);
ram = map_mem_in_cams(ram, CONFIG_LOWMEM_CAM_NUM, true);
linear_sz = min_t(unsigned long, ram, SZ_512M);
@@ -332,6 +344,10 @@ static unsigned long __init kaslr_choose_location(void *dt_ptr, phys_addr_t size
/* If the linear size is smaller than 64M, do not randmize */
if (linear_sz < SZ_64M)
return 0;
+#else
+ ram = size;
+ linear_sz = min_t(unsigned long, size, SZ_1G);
+#endif

/* check for a reserved-memory node and record its cell sizes */
regions.reserved_mem = fdt_path_offset(dt_ptr, "/reserved-memory");
@@ -365,6 +381,14 @@ notrace void __init kaslr_early_init(void *dt_ptr, phys_addr_t size)
unsigned long offset;
unsigned long kernel_sz;

+ if (IS_ENABLED(CONFIG_PPC64)) {
+ if (__run_at_load == 1)
+ return;
+
+ /* Get the first memblock size */
+ early_get_first_memblock_info(dt_ptr, &size);
+ }
+
kernel_sz = (unsigned long)_end - (unsigned long)_stext;

offset = kaslr_choose_location(dt_ptr, size, kernel_sz);
@@ -374,14 +398,19 @@ notrace void __init kaslr_early_init(void *dt_ptr, phys_addr_t size)
kernstart_virt_addr += offset;
kernstart_addr += offset;

- is_second_reloc = 1;
+ if (IS_ENABLED(CONFIG_PPC32)) {
+ is_second_reloc = 1;

- if (offset >= SZ_64M) {
- tlb_virt = round_down(kernstart_virt_addr, SZ_64M);
- tlb_phys = round_down(kernstart_addr, SZ_64M);
+ if (offset >= SZ_64M) {
+ tlb_virt = round_down(kernstart_virt_addr, SZ_64M);
+ tlb_phys = round_down(kernstart_addr, SZ_64M);

- /* Create kernel map to relocate in */
- create_kaslr_tlb_entry(1, tlb_virt, tlb_phys);
+ /* Create kernel map to relocate in */
+ create_kaslr_tlb_entry(1, tlb_virt, tlb_phys);
+ }
+ } else {
+ __kaslr_offset = kernstart_virt_addr - KERNELBASE;
+ __run_at_load = 1;
}

/* Copy the kernel to it's new location and run */
--
2.17.2

2020-04-13 10:47:22

by Jason Yan

[permalink] [raw]
Subject: Re: [PATCH v5 0/6] implement KASLR for powerpc/fsl_booke/64

ping...

?? 2020/3/30 10:20, Jason Yan д??:
> This is a try to implement KASLR for Freescale BookE64 which is based on
> my earlier implementation for Freescale BookE32:
> https://patchwork.ozlabs.org/project/linuxppc-dev/list/?series=131718&state=*
>
> The implementation for Freescale BookE64 is similar as BookE32. One
> difference is that Freescale BookE64 set up a TLB mapping of 1G during
> booting. Another difference is that ppc64 needs the kernel to be
> 64K-aligned. So we can randomize the kernel in this 1G mapping and make
> it 64K-aligned. This can save some code to creat another TLB map at
> early boot. The disadvantage is that we only have about 1G/64K = 16384
> slots to put the kernel in.
>
> KERNELBASE
>
> 64K |--> kernel <--|
> | | |
> +--+--+--+ +--+--+--+--+--+--+--+--+--+ +--+--+
> | | | |....| | | | | | | | | |....| | |
> +--+--+--+ +--+--+--+--+--+--+--+--+--+ +--+--+
> | | 1G
> |-----> offset <-----|
>
> kernstart_virt_addr
>
> I'm not sure if the slot numbers is enough or the design has any
> defects. If you have some better ideas, I would be happy to hear that.
>
> Thank you all.
>
> v4->v5:
> Fix "-Werror=maybe-uninitialized" compile error.
> Fix typo "similar as" -> "similar to".
> v3->v4:
> Do not define __kaslr_offset as a fixed symbol. Reference __run_at_load and
> __kaslr_offset by symbol instead of magic offsets.
> Use IS_ENABLED(CONFIG_PPC32) instead of #ifdef CONFIG_PPC32.
> Change kaslr-booke32 to kaslr-booke in index.rst
> Switch some instructions to 64-bit.
> v2->v3:
> Fix build error when KASLR is disabled.
> v1->v2:
> Add __kaslr_offset for the secondary cpu boot up.
>
> Jason Yan (6):
> powerpc/fsl_booke/kaslr: refactor kaslr_legal_offset() and
> kaslr_early_init()
> powerpc/fsl_booke/64: introduce reloc_kernel_entry() helper
> powerpc/fsl_booke/64: implement KASLR for fsl_booke64
> powerpc/fsl_booke/64: do not clear the BSS for the second pass
> powerpc/fsl_booke/64: clear the original kernel if randomized
> powerpc/fsl_booke/kaslr: rename kaslr-booke32.rst to kaslr-booke.rst
> and add 64bit part
>
> Documentation/powerpc/index.rst | 2 +-
> .../{kaslr-booke32.rst => kaslr-booke.rst} | 35 ++++++-
> arch/powerpc/Kconfig | 2 +-
> arch/powerpc/kernel/exceptions-64e.S | 23 +++++
> arch/powerpc/kernel/head_64.S | 13 +++
> arch/powerpc/kernel/setup_64.c | 3 +
> arch/powerpc/mm/mmu_decl.h | 23 +++--
> arch/powerpc/mm/nohash/kaslr_booke.c | 91 +++++++++++++------
> 8 files changed, 147 insertions(+), 45 deletions(-)
> rename Documentation/powerpc/{kaslr-booke32.rst => kaslr-booke.rst} (59%)
>

2020-04-16 01:04:38

by Crystal Wood

[permalink] [raw]
Subject: Re: [PATCH v5 0/6] implement KASLR for powerpc/fsl_booke/64

On Mon, 2020-03-30 at 10:20 +0800, Jason Yan wrote:
> This is a try to implement KASLR for Freescale BookE64 which is based on
> my earlier implementation for Freescale BookE32:
>
https://patchwork.ozlabs.org/project/linuxppc-dev/list/?series=131718&state=*
>
> The implementation for Freescale BookE64 is similar as BookE32. One
> difference is that Freescale BookE64 set up a TLB mapping of 1G during
> booting. Another difference is that ppc64 needs the kernel to be
> 64K-aligned. So we can randomize the kernel in this 1G mapping and make
> it 64K-aligned. This can save some code to creat another TLB map at
> early boot. The disadvantage is that we only have about 1G/64K = 16384
> slots to put the kernel in.
>
> KERNELBASE
>
> 64K |--> kernel <--|
> | | |
> +--+--+--+ +--+--+--+--+--+--+--+--+--+ +--+--+
> | | | |....| | | | | | | | | |....| | |
> +--+--+--+ +--+--+--+--+--+--+--+--+--+ +--+--+
> | | 1G
> |-----> offset <-----|
>
> kernstart_virt_addr
>
> I'm not sure if the slot numbers is enough or the design has any
> defects. If you have some better ideas, I would be happy to hear that.
>
> Thank you all.
>
> v4->v5:
> Fix "-Werror=maybe-uninitialized" compile error.
> Fix typo "similar as" -> "similar to".
> v3->v4:
> Do not define __kaslr_offset as a fixed symbol. Reference __run_at_load
> and
> __kaslr_offset by symbol instead of magic offsets.
> Use IS_ENABLED(CONFIG_PPC32) instead of #ifdef CONFIG_PPC32.
> Change kaslr-booke32 to kaslr-booke in index.rst
> Switch some instructions to 64-bit.
> v2->v3:
> Fix build error when KASLR is disabled.
> v1->v2:
> Add __kaslr_offset for the secondary cpu boot up.
>
> Jason Yan (6):
> powerpc/fsl_booke/kaslr: refactor kaslr_legal_offset() and
> kaslr_early_init()
> powerpc/fsl_booke/64: introduce reloc_kernel_entry() helper
> powerpc/fsl_booke/64: implement KASLR for fsl_booke64
> powerpc/fsl_booke/64: do not clear the BSS for the second pass
> powerpc/fsl_booke/64: clear the original kernel if randomized
> powerpc/fsl_booke/kaslr: rename kaslr-booke32.rst to kaslr-booke.rst
> and add 64bit part
>
> Documentation/powerpc/index.rst | 2 +-
> .../{kaslr-booke32.rst => kaslr-booke.rst} | 35 ++++++-
> arch/powerpc/Kconfig | 2 +-
> arch/powerpc/kernel/exceptions-64e.S | 23 +++++
> arch/powerpc/kernel/head_64.S | 13 +++
> arch/powerpc/kernel/setup_64.c | 3 +
> arch/powerpc/mm/mmu_decl.h | 23 +++--
> arch/powerpc/mm/nohash/kaslr_booke.c | 91 +++++++++++++------
> 8 files changed, 147 insertions(+), 45 deletions(-)
> rename Documentation/powerpc/{kaslr-booke32.rst => kaslr-booke.rst} (59%)
>

Acked-by: Scott Wood <[email protected]>

-Scott


2020-04-23 01:16:06

by Jason Yan

[permalink] [raw]
Subject: Re: [PATCH v5 0/6] implement KASLR for powerpc/fsl_booke/64

Hi Michael,

What's the status of this series?

Thanks,
Jason

?? 2020/3/30 10:20, Jason Yan д??:
> This is a try to implement KASLR for Freescale BookE64 which is based on
> my earlier implementation for Freescale BookE32:
> https://patchwork.ozlabs.org/project/linuxppc-dev/list/?series=131718&state=*
>
> The implementation for Freescale BookE64 is similar as BookE32. One
> difference is that Freescale BookE64 set up a TLB mapping of 1G during
> booting. Another difference is that ppc64 needs the kernel to be
> 64K-aligned. So we can randomize the kernel in this 1G mapping and make
> it 64K-aligned. This can save some code to creat another TLB map at
> early boot. The disadvantage is that we only have about 1G/64K = 16384
> slots to put the kernel in.
>
> KERNELBASE
>
> 64K |--> kernel <--|
> | | |
> +--+--+--+ +--+--+--+--+--+--+--+--+--+ +--+--+
> | | | |....| | | | | | | | | |....| | |
> +--+--+--+ +--+--+--+--+--+--+--+--+--+ +--+--+
> | | 1G
> |-----> offset <-----|
>
> kernstart_virt_addr
>
> I'm not sure if the slot numbers is enough or the design has any
> defects. If you have some better ideas, I would be happy to hear that.
>
> Thank you all.
>
> v4->v5:
> Fix "-Werror=maybe-uninitialized" compile error.
> Fix typo "similar as" -> "similar to".
> v3->v4:
> Do not define __kaslr_offset as a fixed symbol. Reference __run_at_load and
> __kaslr_offset by symbol instead of magic offsets.
> Use IS_ENABLED(CONFIG_PPC32) instead of #ifdef CONFIG_PPC32.
> Change kaslr-booke32 to kaslr-booke in index.rst
> Switch some instructions to 64-bit.
> v2->v3:
> Fix build error when KASLR is disabled.
> v1->v2:
> Add __kaslr_offset for the secondary cpu boot up.
>
> Jason Yan (6):
> powerpc/fsl_booke/kaslr: refactor kaslr_legal_offset() and
> kaslr_early_init()
> powerpc/fsl_booke/64: introduce reloc_kernel_entry() helper
> powerpc/fsl_booke/64: implement KASLR for fsl_booke64
> powerpc/fsl_booke/64: do not clear the BSS for the second pass
> powerpc/fsl_booke/64: clear the original kernel if randomized
> powerpc/fsl_booke/kaslr: rename kaslr-booke32.rst to kaslr-booke.rst
> and add 64bit part
>
> Documentation/powerpc/index.rst | 2 +-
> .../{kaslr-booke32.rst => kaslr-booke.rst} | 35 ++++++-
> arch/powerpc/Kconfig | 2 +-
> arch/powerpc/kernel/exceptions-64e.S | 23 +++++
> arch/powerpc/kernel/head_64.S | 13 +++
> arch/powerpc/kernel/setup_64.c | 3 +
> arch/powerpc/mm/mmu_decl.h | 23 +++--
> arch/powerpc/mm/nohash/kaslr_booke.c | 91 +++++++++++++------
> 8 files changed, 147 insertions(+), 45 deletions(-)
> rename Documentation/powerpc/{kaslr-booke32.rst => kaslr-booke.rst} (59%)
>

2020-04-24 17:22:09

by Daniel Axtens

[permalink] [raw]
Subject: Re: [PATCH v5 0/6] implement KASLR for powerpc/fsl_booke/64


> This is a try to implement KASLR for Freescale BookE64 which is based on
> my earlier implementation for Freescale BookE32:
> https://patchwork.ozlabs.org/project/linuxppc-dev/list/?series=131718&state=*
>
> The implementation for Freescale BookE64 is similar as BookE32. One
> difference is that Freescale BookE64 set up a TLB mapping of 1G during
> booting. Another difference is that ppc64 needs the kernel to be
> 64K-aligned. So we can randomize the kernel in this 1G mapping and make
> it 64K-aligned. This can save some code to creat another TLB map at
> early boot. The disadvantage is that we only have about 1G/64K = 16384
> slots to put the kernel in.
>
> KERNELBASE
>
> 64K |--> kernel <--|
> | | |
> +--+--+--+ +--+--+--+--+--+--+--+--+--+ +--+--+
> | | | |....| | | | | | | | | |....| | |
> +--+--+--+ +--+--+--+--+--+--+--+--+--+ +--+--+
> | | 1G
> |-----> offset <-----|
>
> kernstart_virt_addr
>
> I'm not sure if the slot numbers is enough or the design has any
> defects. If you have some better ideas, I would be happy to hear that.
>
> Thank you all.
>
> v4->v5:
> Fix "-Werror=maybe-uninitialized" compile error.
> Fix typo "similar as" -> "similar to".
> v3->v4:
> Do not define __kaslr_offset as a fixed symbol. Reference __run_at_load and
> __kaslr_offset by symbol instead of magic offsets.
> Use IS_ENABLED(CONFIG_PPC32) instead of #ifdef CONFIG_PPC32.
> Change kaslr-booke32 to kaslr-booke in index.rst
> Switch some instructions to 64-bit.
> v2->v3:
> Fix build error when KASLR is disabled.
> v1->v2:
> Add __kaslr_offset for the secondary cpu boot up.
>
> Jason Yan (6):
> powerpc/fsl_booke/kaslr: refactor kaslr_legal_offset() and
> kaslr_early_init()
> powerpc/fsl_booke/64: introduce reloc_kernel_entry() helper
> powerpc/fsl_booke/64: implement KASLR for fsl_booke64
> powerpc/fsl_booke/64: do not clear the BSS for the second pass
> powerpc/fsl_booke/64: clear the original kernel if randomized
> powerpc/fsl_booke/kaslr: rename kaslr-booke32.rst to kaslr-booke.rst
> and add 64bit part
>
> Documentation/powerpc/index.rst | 2 +-
> .../{kaslr-booke32.rst => kaslr-booke.rst} | 35 ++++++-
> arch/powerpc/Kconfig | 2 +-
> arch/powerpc/kernel/exceptions-64e.S | 23 +++++
> arch/powerpc/kernel/head_64.S | 13 +++
> arch/powerpc/kernel/setup_64.c | 3 +
> arch/powerpc/mm/mmu_decl.h | 23 +++--
> arch/powerpc/mm/nohash/kaslr_booke.c | 91 +++++++++++++------
> 8 files changed, 147 insertions(+), 45 deletions(-)
> rename Documentation/powerpc/{kaslr-booke32.rst => kaslr-booke.rst} (59%)
>
> --
> 2.17.2


Attachments:
.config (92.52 kB)

2020-04-26 04:00:36

by Jason Yan

[permalink] [raw]
Subject: Re: [PATCH v5 0/6] implement KASLR for powerpc/fsl_booke/64

Hi Daniel,

Thanks for the test. Can you send me the full log which may contain the
system info such as the following:

-----------------------------------------------------
phys_mem_size = 0x200000000
dcache_bsize = 0x20
icache_bsize = 0x20
cpu_features = 0x00000003008003b6
possible = 0x00000003009003b6
always = 0x00000003008003b4
cpu_user_features = 0xdc008000 0x08000000
mmu_features = 0x000a0010
firmware_features = 0x0000000000000000
physical_start = 0x20000000
-----------------------------------------------------
barrier-nospec: using isync; sync as speculation barrier
Zone ranges:
DMA [mem 0x0000000000000000-0x000000007fffffff]
Normal [mem 0x0000000080000000-0x00000001ffffffff]
Movable zone start for each node
Early memory node ranges
node 0: [mem 0x0000000000000000-0x00000001ffffffff]
Initmem setup node 0 [mem 0x0000000000000000-0x00000001ffffffff]
MMU: Allocated 2112 bytes of context maps for 255 contexts
percpu: Embedded 32 pages/cpu s91736 r0 d39336 u131072
Built 1 zonelists, mobility grouping on. Total pages: 2064384
Kernel command line: console=ttyS0 root=/dev/ram0
printk: log_buf_len individual max cpu contribution: 16384 bytes
printk: log_buf_len total cpu_extra contributions: 376832 bytes
printk: log_buf_len min size: 131072 bytes
printk: log_buf_len: 524288 bytes
printk: early log buf free: 127460(97%)
Dentry cache hash table entries: 1048576 (order: 11, 8388608 bytes, linear)
Inode-cache hash table entries: 524288 (order: 10, 4194304 bytes, linear)
mem auto-init: stack:off, heap alloc:off, heap free:off
Memory: 8135632K/8388608K available (10572K kernel code, 2000K rwdata,
3396K rodata, 4124K init, 358K bss, 252976K reserved, 0K cma-reserved)
rcu: Hierarchical RCU implementation.
rcu: RCU event tracing is enabled.
rcu: RCU calculated value of scheduler-enlistment delay is 25 jiffies.
NR_IRQS: 512, nr_irqs: 512, preallocated irqs: 16
mpic: Setting up MPIC " OpenPIC " version 1.2 at fe0040000, max 24 CPUs
mpic: ISU size: 256, shift: 8, mask: ff
mpic: Initializing for 256 sources
random: get_random_u64 called from .start_kernel+0x724/0x954 with
crng_init=0
clocksource: timebase: mask: 0xffffffffffffffff max_cycles:
0x5c4093a7d1, max_idle_ns: 440795210635 ns
clocksource: timebase mult[2800000] shift[24] registered
Console: colour dummy device 80x25
pid_max: default: 32768 minimum: 301
Mount-cache hash table entries: 16384 (order: 5, 131072 bytes, linear)
Mountpoint-cache hash table entries: 16384 (order: 5, 131072 bytes, linear)
e6500 family performance monitor hardware support registered
rcu: Hierarchical SRCU implementation.
smp: Bringing up secondary CPUs ...
smp: Brought up 1 node, 24 CPUs



?? 2020/4/25 1:17, Daniel Axtens д??:
> Hi Jason,
>
> Apologies for the delay in testing.
>
> I'm seeing this problem when I try to boot on a t4240rdb:
>
> random: get_random_u64 called from .start_kernel+0x734/0x964 with crng_init=0 [8/973]
> clocksource: timebase: mask: 0xffffffffffffffff max_cycles: 0xa9210e89c, max_idle_ns: 440795203878 ns
> clocksource: timebase mult[15d17460] shift[24] registered
> Console: colour dummy device 80x25
> pid_max: default: 32768 minimum: 301
> Mount-cache hash table entries: 16384 (order: 5, 131072 bytes, linear)
> Mountpoint-cache hash table entries: 16384 (order: 5, 131072 bytes, linear)
> e6500 family performance monitor hardware support registered
> rcu: Hierarchical SRCU implementation.
> smp: Bringing up secondary CPUs ...
> Processor 2 is stuck.
> Processor 3 is stuck.
> Processor 4 is stuck.
> Processor 5 is stuck.
> Processor 6 is stuck.
> Processor 7 is stuck.
> Processor 8 is stuck.
> Processor 9 is stuck.
> Processor 10 is stuck.
> Processor 11 is stuck.
> Processor 12 is stuck.
> Processor 13 is stuck.
> Processor 14 is stuck.
> ...
> Processor 22 is stuck.
> Processor 23 is stuck.
> smp: Brought up 1 node, 2 CPUs
> Using standard scheduler topology
> devtmpfs: initialized
> clocksource: jiffies: mask: 0xffffffff max_cycles: 0xffffffff, max_idle_ns: 7645041785100000 ns
> futex hash table entries: 8192 (order: 7, 524288 bytes, linear)
> NET: Registered protocol family 16
> audit: initializing netlink subsys (disabled)
> audit: type=2000 audit(108.032:1): state=initialized audit_enabled=0 res=1
> Machine: fsl,T4240RDB
> SoC family: QorIQ T4240
> SoC ID: svr:0x82480020, Revision: 2.0
> ... boot continues ...
>
>
> If I boot with nokaslr, all the CPUs come up with no issue.
>
> This is on top of powerpc/merge at
> 8299da600ad05b8aa0f15ec0f5f03bd40e37d6f0. If you'd like me to test any
> debug patches I can do that.
>
> I've attached my .config.
>
> Regards,
> Daniel
>
>
>
>> This is a try to implement KASLR for Freescale BookE64 which is based on
>> my earlier implementation for Freescale BookE32:
>> https://patchwork.ozlabs.org/project/linuxppc-dev/list/?series=131718&state=*
>>
>> The implementation for Freescale BookE64 is similar as BookE32. One
>> difference is that Freescale BookE64 set up a TLB mapping of 1G during
>> booting. Another difference is that ppc64 needs the kernel to be
>> 64K-aligned. So we can randomize the kernel in this 1G mapping and make
>> it 64K-aligned. This can save some code to creat another TLB map at
>> early boot. The disadvantage is that we only have about 1G/64K = 16384
>> slots to put the kernel in.
>>
>> KERNELBASE
>>
>> 64K |--> kernel <--|
>> | | |
>> +--+--+--+ +--+--+--+--+--+--+--+--+--+ +--+--+
>> | | | |....| | | | | | | | | |....| | |
>> +--+--+--+ +--+--+--+--+--+--+--+--+--+ +--+--+
>> | | 1G
>> |-----> offset <-----|
>>
>> kernstart_virt_addr
>>
>> I'm not sure if the slot numbers is enough or the design has any
>> defects. If you have some better ideas, I would be happy to hear that.
>>
>> Thank you all.
>>
>> v4->v5:
>> Fix "-Werror=maybe-uninitialized" compile error.
>> Fix typo "similar as" -> "similar to".
>> v3->v4:
>> Do not define __kaslr_offset as a fixed symbol. Reference __run_at_load and
>> __kaslr_offset by symbol instead of magic offsets.
>> Use IS_ENABLED(CONFIG_PPC32) instead of #ifdef CONFIG_PPC32.
>> Change kaslr-booke32 to kaslr-booke in index.rst
>> Switch some instructions to 64-bit.
>> v2->v3:
>> Fix build error when KASLR is disabled.
>> v1->v2:
>> Add __kaslr_offset for the secondary cpu boot up.
>>
>> Jason Yan (6):
>> powerpc/fsl_booke/kaslr: refactor kaslr_legal_offset() and
>> kaslr_early_init()
>> powerpc/fsl_booke/64: introduce reloc_kernel_entry() helper
>> powerpc/fsl_booke/64: implement KASLR for fsl_booke64
>> powerpc/fsl_booke/64: do not clear the BSS for the second pass
>> powerpc/fsl_booke/64: clear the original kernel if randomized
>> powerpc/fsl_booke/kaslr: rename kaslr-booke32.rst to kaslr-booke.rst
>> and add 64bit part
>>
>> Documentation/powerpc/index.rst | 2 +-
>> .../{kaslr-booke32.rst => kaslr-booke.rst} | 35 ++++++-
>> arch/powerpc/Kconfig | 2 +-
>> arch/powerpc/kernel/exceptions-64e.S | 23 +++++
>> arch/powerpc/kernel/head_64.S | 13 +++
>> arch/powerpc/kernel/setup_64.c | 3 +
>> arch/powerpc/mm/mmu_decl.h | 23 +++--
>> arch/powerpc/mm/nohash/kaslr_booke.c | 91 +++++++++++++------
>> 8 files changed, 147 insertions(+), 45 deletions(-)
>> rename Documentation/powerpc/{kaslr-booke32.rst => kaslr-booke.rst} (59%)
>>
>> --
>> 2.17.2

2020-04-27 13:38:43

by Daniel Axtens

[permalink] [raw]
Subject: Re: [PATCH v5 0/6] implement KASLR for powerpc/fsl_booke/64



>
> -----------------------------------------------------
> phys_mem_size = 0x200000000
> dcache_bsize = 0x20
> icache_bsize = 0x20
> cpu_features = 0x00000003008003b6
> possible = 0x00000003009003b6
> always = 0x00000003008003b4
> cpu_user_features = 0xdc008000 0x08000000
> mmu_features = 0x000a0010
> firmware_features = 0x0000000000000000
> physical_start = 0x20000000
> -----------------------------------------------------
> barrier-nospec: using isync; sync as speculation barrier
> Zone ranges:
> DMA [mem 0x0000000000000000-0x000000007fffffff]
> Normal [mem 0x0000000080000000-0x00000001ffffffff]
> Movable zone start for each node
> Early memory node ranges
> node 0: [mem 0x0000000000000000-0x00000001ffffffff]
> Initmem setup node 0 [mem 0x0000000000000000-0x00000001ffffffff]
> MMU: Allocated 2112 bytes of context maps for 255 contexts
> percpu: Embedded 32 pages/cpu s91736 r0 d39336 u131072
> Built 1 zonelists, mobility grouping on. Total pages: 2064384
> Kernel command line: console=ttyS0 root=/dev/ram0
> printk: log_buf_len individual max cpu contribution: 16384 bytes
> printk: log_buf_len total cpu_extra contributions: 376832 bytes
> printk: log_buf_len min size: 131072 bytes
> printk: log_buf_len: 524288 bytes
> printk: early log buf free: 127460(97%)
> Dentry cache hash table entries: 1048576 (order: 11, 8388608 bytes, linear)
> Inode-cache hash table entries: 524288 (order: 10, 4194304 bytes, linear)
> mem auto-init: stack:off, heap alloc:off, heap free:off
> Memory: 8135632K/8388608K available (10572K kernel code, 2000K rwdata,
> 3396K rodata, 4124K init, 358K bss, 252976K reserved, 0K cma-reserved)
> rcu: Hierarchical RCU implementation.
> rcu: RCU event tracing is enabled.
> rcu: RCU calculated value of scheduler-enlistment delay is 25 jiffies.
> NR_IRQS: 512, nr_irqs: 512, preallocated irqs: 16
> mpic: Setting up MPIC " OpenPIC " version 1.2 at fe0040000, max 24 CPUs
> mpic: ISU size: 256, shift: 8, mask: ff
> mpic: Initializing for 256 sources
> random: get_random_u64 called from .start_kernel+0x724/0x954 with
> crng_init=0
> clocksource: timebase: mask: 0xffffffffffffffff max_cycles:
> 0x5c4093a7d1, max_idle_ns: 440795210635 ns
> clocksource: timebase mult[2800000] shift[24] registered
> Console: colour dummy device 80x25
> pid_max: default: 32768 minimum: 301
> Mount-cache hash table entries: 16384 (order: 5, 131072 bytes, linear)
> Mountpoint-cache hash table entries: 16384 (order: 5, 131072 bytes, linear)
> e6500 family performance monitor hardware support registered
> rcu: Hierarchical SRCU implementation.
> smp: Bringing up secondary CPUs ...
> smp: Brought up 1 node, 24 CPUs
>
>
>
> 在 2020/4/25 1:17, Daniel Axtens 写道:
>> Hi Jason,
>>
>> Apologies for the delay in testing.
>>
>> I'm seeing this problem when I try to boot on a t4240rdb:
>>
>> random: get_random_u64 called from .start_kernel+0x734/0x964 with crng_init=0 [8/973]
>> clocksource: timebase: mask: 0xffffffffffffffff max_cycles: 0xa9210e89c, max_idle_ns: 440795203878 ns
>> clocksource: timebase mult[15d17460] shift[24] registered
>> Console: colour dummy device 80x25
>> pid_max: default: 32768 minimum: 301
>> Mount-cache hash table entries: 16384 (order: 5, 131072 bytes, linear)
>> Mountpoint-cache hash table entries: 16384 (order: 5, 131072 bytes, linear)
>> e6500 family performance monitor hardware support registered
>> rcu: Hierarchical SRCU implementation.
>> smp: Bringing up secondary CPUs ...
>> Processor 2 is stuck.
>> Processor 3 is stuck.
>> Processor 4 is stuck.
>> Processor 5 is stuck.
>> Processor 6 is stuck.
>> Processor 7 is stuck.
>> Processor 8 is stuck.
>> Processor 9 is stuck.
>> Processor 10 is stuck.
>> Processor 11 is stuck.
>> Processor 12 is stuck.
>> Processor 13 is stuck.
>> Processor 14 is stuck.
>> ...
>> Processor 22 is stuck.
>> Processor 23 is stuck.
>> smp: Brought up 1 node, 2 CPUs
>> Using standard scheduler topology
>> devtmpfs: initialized
>> clocksource: jiffies: mask: 0xffffffff max_cycles: 0xffffffff, max_idle_ns: 7645041785100000 ns
>> futex hash table entries: 8192 (order: 7, 524288 bytes, linear)
>> NET: Registered protocol family 16
>> audit: initializing netlink subsys (disabled)
>> audit: type=2000 audit(108.032:1): state=initialized audit_enabled=0 res=1
>> Machine: fsl,T4240RDB
>> SoC family: QorIQ T4240
>> SoC ID: svr:0x82480020, Revision: 2.0
>> ... boot continues ...
>>
>>
>> If I boot with nokaslr, all the CPUs come up with no issue.
>>
>> This is on top of powerpc/merge at
>> 8299da600ad05b8aa0f15ec0f5f03bd40e37d6f0. If you'd like me to test any
>> debug patches I can do that.
>>
>> I've attached my .config.
>>
>> Regards,
>> Daniel
>>
>>
>>
>>> This is a try to implement KASLR for Freescale BookE64 which is based on
>>> my earlier implementation for Freescale BookE32:
>>> https://patchwork.ozlabs.org/project/linuxppc-dev/list/?series=131718&state=*
>>>
>>> The implementation for Freescale BookE64 is similar as BookE32. One
>>> difference is that Freescale BookE64 set up a TLB mapping of 1G during
>>> booting. Another difference is that ppc64 needs the kernel to be
>>> 64K-aligned. So we can randomize the kernel in this 1G mapping and make
>>> it 64K-aligned. This can save some code to creat another TLB map at
>>> early boot. The disadvantage is that we only have about 1G/64K = 16384
>>> slots to put the kernel in.
>>>
>>> KERNELBASE
>>>
>>> 64K |--> kernel <--|
>>> | | |
>>> +--+--+--+ +--+--+--+--+--+--+--+--+--+ +--+--+
>>> | | | |....| | | | | | | | | |....| | |
>>> +--+--+--+ +--+--+--+--+--+--+--+--+--+ +--+--+
>>> | | 1G
>>> |-----> offset <-----|
>>>
>>> kernstart_virt_addr
>>>
>>> I'm not sure if the slot numbers is enough or the design has any
>>> defects. If you have some better ideas, I would be happy to hear that.
>>>
>>> Thank you all.
>>>
>>> v4->v5:
>>> Fix "-Werror=maybe-uninitialized" compile error.
>>> Fix typo "similar as" -> "similar to".
>>> v3->v4:
>>> Do not define __kaslr_offset as a fixed symbol. Reference __run_at_load and
>>> __kaslr_offset by symbol instead of magic offsets.
>>> Use IS_ENABLED(CONFIG_PPC32) instead of #ifdef CONFIG_PPC32.
>>> Change kaslr-booke32 to kaslr-booke in index.rst
>>> Switch some instructions to 64-bit.
>>> v2->v3:
>>> Fix build error when KASLR is disabled.
>>> v1->v2:
>>> Add __kaslr_offset for the secondary cpu boot up.
>>>
>>> Jason Yan (6):
>>> powerpc/fsl_booke/kaslr: refactor kaslr_legal_offset() and
>>> kaslr_early_init()
>>> powerpc/fsl_booke/64: introduce reloc_kernel_entry() helper
>>> powerpc/fsl_booke/64: implement KASLR for fsl_booke64
>>> powerpc/fsl_booke/64: do not clear the BSS for the second pass
>>> powerpc/fsl_booke/64: clear the original kernel if randomized
>>> powerpc/fsl_booke/kaslr: rename kaslr-booke32.rst to kaslr-booke.rst
>>> and add 64bit part
>>>
>>> Documentation/powerpc/index.rst | 2 +-
>>> .../{kaslr-booke32.rst => kaslr-booke.rst} | 35 ++++++-
>>> arch/powerpc/Kconfig | 2 +-
>>> arch/powerpc/kernel/exceptions-64e.S | 23 +++++
>>> arch/powerpc/kernel/head_64.S | 13 +++
>>> arch/powerpc/kernel/setup_64.c | 3 +
>>> arch/powerpc/mm/mmu_decl.h | 23 +++--
>>> arch/powerpc/mm/nohash/kaslr_booke.c | 91 +++++++++++++------
>>> 8 files changed, 147 insertions(+), 45 deletions(-)
>>> rename Documentation/powerpc/{kaslr-booke32.rst => kaslr-booke.rst} (59%)
>>>
>>> --
>>> 2.17.2


Attachments:
fail-log (39.49 kB)