* Re: [RFC PATCH 7/8] powerpc/purgatory: drop .machine specifier
From: Michael Ellerman @ 2021-03-19 10:59 UTC (permalink / raw)
To: Nicholas Piggin, Daniel Axtens, Segher Boessenkool
Cc: linuxppc-dev, llvmlinux
In-Reply-To: <1616119361.tyoejtbh8j.astroid@bobo.none>
Nicholas Piggin <npiggin@gmail.com> writes:
> Excerpts from Segher Boessenkool's message of February 26, 2021 1:58 am:
>> On Thu, Feb 25, 2021 at 02:10:05PM +1100, Daniel Axtens wrote:
>>> It's ignored by future versions of llvm's integrated assembler (by not -11).
>>> I'm not sure what it does for us in gas.
>>
>> It enables all insns that exist on 620 (the first 64-bit PowerPC CPU).
>
> Same question for this, why do we have it at all?
I sent a patch to drop it.
https://patchwork.ozlabs.org/project/linuxppc-dev/patch/20210315034159.315675-1-mpe@ellerman.id.au/
cheers
^ permalink raw reply
* [PATCH 00/10] Convert signal32 to user read access by block
From: Christophe Leroy @ 2021-03-19 11:06 UTC (permalink / raw)
To: Benjamin Herrenschmidt, Paul Mackerras, Michael Ellerman, cmr
Cc: linux-arch, linuxppc-dev, linux-kernel
Similarly to the work done earlier with writes, this series
converts signal32 to using user_read_access_begin/end and
unsafe_get_user() and friends.
Applies on to of the signal64 series, ie on merge-test (ca6e327fefb2)
Christophe Leroy (10):
signal: Add unsafe_get_compat_sigset()
powerpc/uaccess: Also perform 64 bits copies in
unsafe_copy_from_user() on ppc32
powerpc/signal: Add unsafe_copy_ck{fpr/vsx}_from_user
powerpc/signal32: Rename save_user_regs_unsafe() and
save_general_regs_unsafe()
powerpc/signal32: Remove ifdefery in middle of if/else in sigreturn()
powerpc/signal32: Perform access_ok() inside restore_user_regs()
powerpc/signal32: Reorder user reads in restore_tm_user_regs()
powerpc/signal32: Convert restore_[tm]_user_regs() to user access
block
powerpc/signal32: Convert do_setcontext[_tm]() to user access block
powerpc/signal32: Simplify logging in sigreturn()
arch/powerpc/include/asm/ptrace.h | 2 +-
arch/powerpc/include/asm/uaccess.h | 6 +-
arch/powerpc/kernel/signal.h | 22 +++
arch/powerpc/kernel/signal_32.c | 251 ++++++++++++++++-------------
include/linux/compat.h | 35 ++++
include/linux/uaccess.h | 1 +
6 files changed, 205 insertions(+), 112 deletions(-)
--
2.25.0
^ permalink raw reply
* [PATCH 02/10] powerpc/uaccess: Also perform 64 bits copies in unsafe_copy_from_user() on ppc32
From: Christophe Leroy @ 2021-03-19 11:06 UTC (permalink / raw)
To: Benjamin Herrenschmidt, Paul Mackerras, Michael Ellerman, cmr
Cc: linuxppc-dev, linux-kernel
In-Reply-To: <cover.1616151715.git.christophe.leroy@csgroup.eu>
Similarly to commit 5cf773fc8f37 ("powerpc/uaccess: Also perform
64 bits copies in unsafe_copy_to_user() on ppc32")
ppc32 has an efficiant 64 bits unsafe_get_user(), so also use it in
order to unroll loops more.
Signed-off-by: Christophe Leroy <christophe.leroy@csgroup.eu>
---
arch/powerpc/include/asm/uaccess.h | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/arch/powerpc/include/asm/uaccess.h b/arch/powerpc/include/asm/uaccess.h
index 77d837b16e4d..a4e791bcd3fe 100644
--- a/arch/powerpc/include/asm/uaccess.h
+++ b/arch/powerpc/include/asm/uaccess.h
@@ -432,9 +432,9 @@ do { \
size_t _len = (l); \
int _i; \
\
- for (_i = 0; _i < (_len & ~(sizeof(long) - 1)); _i += sizeof(long)) \
- unsafe_get_user(*(long *)(_dst + _i), (long __user *)(_src + _i), e); \
- if (IS_ENABLED(CONFIG_PPC64) && (_len & 4)) { \
+ for (_i = 0; _i < (_len & ~(sizeof(u64) - 1)); _i += sizeof(u64)) \
+ unsafe_get_user(*(u64 *)(_dst + _i), (u64 __user *)(_src + _i), e); \
+ if (_len & 4) { \
unsafe_get_user(*(u32 *)(_dst + _i), (u32 __user *)(_src + _i), e); \
_i += 4; \
} \
--
2.25.0
^ permalink raw reply related
* [PATCH 03/10] powerpc/signal: Add unsafe_copy_ck{fpr/vsx}_from_user
From: Christophe Leroy @ 2021-03-19 11:06 UTC (permalink / raw)
To: Benjamin Herrenschmidt, Paul Mackerras, Michael Ellerman, cmr
Cc: linuxppc-dev, linux-kernel
In-Reply-To: <cover.1616151715.git.christophe.leroy@csgroup.eu>
Add unsafe_copy_ckfpr_from_user() and unsafe_copy_ckvsx_from_user()
Signed-off-by: Christophe Leroy <christophe.leroy@csgroup.eu>
---
arch/powerpc/kernel/signal.h | 20 ++++++++++++++++++++
1 file changed, 20 insertions(+)
diff --git a/arch/powerpc/kernel/signal.h b/arch/powerpc/kernel/signal.h
index 1393876f3814..a5152ff3c52f 100644
--- a/arch/powerpc/kernel/signal.h
+++ b/arch/powerpc/kernel/signal.h
@@ -100,6 +100,26 @@ unsigned long copy_ckfpr_from_user(struct task_struct *task, void __user *from);
unsafe_put_user(__t->thread.ckfp_state.fpr[i][TS_VSRLOWOFFSET], \
&buf[i], label);\
} while (0)
+
+#define unsafe_copy_ckfpr_from_user(task, from, label) do { \
+ struct task_struct *__t = task; \
+ u64 __user *buf = (u64 __user *)from; \
+ int i; \
+ \
+ for (i = 0; i < ELF_NFPREG - 1 ; i++) \
+ unsafe_get_user(__t->thread.TS_CKFPR(i), &buf[i], label);\
+ unsafe_get_user(__t->thread.ckfp_state.fpscr, &buf[i], failed); \
+} while (0)
+
+#define unsafe_copy_ckvsx_from_user(task, from, label) do { \
+ struct task_struct *__t = task; \
+ u64 __user *buf = (u64 __user *)from; \
+ int i; \
+ \
+ for (i = 0; i < ELF_NVSRHALFREG ; i++) \
+ unsafe_get_user(__t->thread.ckfp_state.fpr[i][TS_VSRLOWOFFSET], \
+ &buf[i], label); \
+} while (0)
#endif
#elif defined(CONFIG_PPC_FPU_REGS)
--
2.25.0
^ permalink raw reply related
* [PATCH 01/10] signal: Add unsafe_get_compat_sigset()
From: Christophe Leroy @ 2021-03-19 11:06 UTC (permalink / raw)
To: Benjamin Herrenschmidt, Paul Mackerras, Michael Ellerman, cmr
Cc: linux-arch, linuxppc-dev, linux-kernel
In-Reply-To: <cover.1616151715.git.christophe.leroy@csgroup.eu>
In the same way as commit 14026b94ccfe ("signal: Add
unsafe_put_compat_sigset()"), this time add
unsafe_get_compat_sigset() macro which is the 'unsafe'
version of get_compat_sigset()
For the bigendian, use unsafe_get_user() directly
to avoid intermediate copy through the stack.
For the littleendian, use a straight unsafe_copy_from_user().
This commit adds the generic fallback for unsafe_copy_from_user().
Architectures wanting to use unsafe_get_compat_sigset() have to
make sure they have their own unsafe_copy_from_user().
Signed-off-by: Christophe Leroy <christophe.leroy@csgroup.eu>
---
include/linux/compat.h | 35 +++++++++++++++++++++++++++++++++++
include/linux/uaccess.h | 1 +
2 files changed, 36 insertions(+)
diff --git a/include/linux/compat.h b/include/linux/compat.h
index 6e65be753603..5112c3e35782 100644
--- a/include/linux/compat.h
+++ b/include/linux/compat.h
@@ -465,6 +465,34 @@ put_compat_sigset(compat_sigset_t __user *compat, const sigset_t *set,
unsafe_put_user(__s->sig[0], &__c->sig[0], label); \
} \
} while (0)
+
+#define unsafe_get_compat_sigset(set, compat, label) do { \
+ const compat_sigset_t __user *__c = compat; \
+ compat_sigset_word hi, lo; \
+ sigset_t *__s = set; \
+ \
+ switch (_NSIG_WORDS) { \
+ case 4: \
+ unsafe_get_user(lo, &__c->sig[7], label); \
+ unsafe_get_user(hi, &__c->sig[6], label); \
+ __s->sig[3] = hi | (((long)lo) << 32); \
+ fallthrough; \
+ case 3: \
+ unsafe_get_user(lo, &__c->sig[5], label); \
+ unsafe_get_user(hi, &__c->sig[4], label); \
+ __s->sig[2] = hi | (((long)lo) << 32); \
+ fallthrough; \
+ case 2: \
+ unsafe_get_user(lo, &__c->sig[3], label); \
+ unsafe_get_user(hi, &__c->sig[2], label); \
+ __s->sig[1] = hi | (((long)lo) << 32); \
+ fallthrough; \
+ case 1: \
+ unsafe_get_user(lo, &__c->sig[1], label); \
+ unsafe_get_user(hi, &__c->sig[0], label); \
+ __s->sig[0] = hi | (((long)lo) << 32); \
+ } \
+} while (0)
#else
#define unsafe_put_compat_sigset(compat, set, label) do { \
compat_sigset_t __user *__c = compat; \
@@ -472,6 +500,13 @@ put_compat_sigset(compat_sigset_t __user *compat, const sigset_t *set,
\
unsafe_copy_to_user(__c, __s, sizeof(*__c), label); \
} while (0)
+
+#define unsafe_get_compat_sigset(set, compat, label) do { \
+ const compat_sigset_t __user *__c = compat; \
+ sigset_t *__s = set; \
+ \
+ unsafe_copy_from_user(__s, __c, sizeof(*__c), label); \
+} while (0)
#endif
extern int compat_ptrace_request(struct task_struct *child,
diff --git a/include/linux/uaccess.h b/include/linux/uaccess.h
index c7c6e8b8344d..c05e903cef02 100644
--- a/include/linux/uaccess.h
+++ b/include/linux/uaccess.h
@@ -397,6 +397,7 @@ long strnlen_user_nofault(const void __user *unsafe_addr, long count);
#define unsafe_get_user(x,p,e) unsafe_op_wrap(__get_user(x,p),e)
#define unsafe_put_user(x,p,e) unsafe_op_wrap(__put_user(x,p),e)
#define unsafe_copy_to_user(d,s,l,e) unsafe_op_wrap(__copy_to_user(d,s,l),e)
+#define unsafe_copy_from_user(d,s,l,e) unsafe_op_wrap(__copy_from_user(d,s,l),e)
static inline unsigned long user_access_save(void) { return 0UL; }
static inline void user_access_restore(unsigned long flags) { }
#endif
--
2.25.0
^ permalink raw reply related
* [PATCH 04/10] powerpc/signal32: Rename save_user_regs_unsafe() and save_general_regs_unsafe()
From: Christophe Leroy @ 2021-03-19 11:06 UTC (permalink / raw)
To: Benjamin Herrenschmidt, Paul Mackerras, Michael Ellerman, cmr
Cc: linuxppc-dev, linux-kernel
In-Reply-To: <cover.1616151715.git.christophe.leroy@csgroup.eu>
Convention is to prefix functions with __unsafe_ instead of
suffixing it with _unsafe.
Rename save_user_regs_unsafe() and save_general_regs_unsafe()
accordingly, that is respectively __unsafe_save_general_regs() and
__unsafe_save_user_regs().
Suggested-by: Christopher M. Riedl <cmr@codefail.de>
Signed-off-by: Christophe Leroy <christophe.leroy@csgroup.eu>
---
arch/powerpc/kernel/signal_32.c | 16 ++++++++--------
1 file changed, 8 insertions(+), 8 deletions(-)
diff --git a/arch/powerpc/kernel/signal_32.c b/arch/powerpc/kernel/signal_32.c
index c505b444a613..3b78748d6d85 100644
--- a/arch/powerpc/kernel/signal_32.c
+++ b/arch/powerpc/kernel/signal_32.c
@@ -94,7 +94,7 @@ static inline int get_sigset_t(sigset_t *set,
#define from_user_ptr(p) compat_ptr(p)
static __always_inline int
-save_general_regs_unsafe(struct pt_regs *regs, struct mcontext __user *frame)
+__unsafe_save_general_regs(struct pt_regs *regs, struct mcontext __user *frame)
{
elf_greg_t64 *gregs = (elf_greg_t64 *)regs;
int val, i;
@@ -151,7 +151,7 @@ static inline int get_sigset_t(sigset_t *set, const sigset_t __user *uset)
#define from_user_ptr(p) ((void __user *)(p))
static __always_inline int
-save_general_regs_unsafe(struct pt_regs *regs, struct mcontext __user *frame)
+__unsafe_save_general_regs(struct pt_regs *regs, struct mcontext __user *frame)
{
WARN_ON(!FULL_REGS(regs));
unsafe_copy_to_user(&frame->mc_gregs, regs, GP_REGS_SIZE, failed);
@@ -177,7 +177,7 @@ static inline int restore_general_regs(struct pt_regs *regs,
#endif
#define unsafe_save_general_regs(regs, frame, label) do { \
- if (save_general_regs_unsafe(regs, frame)) \
+ if (__unsafe_save_general_regs(regs, frame)) \
goto label; \
} while (0)
@@ -260,8 +260,8 @@ static void prepare_save_user_regs(int ctx_has_vsx_region)
#endif
}
-static int save_user_regs_unsafe(struct pt_regs *regs, struct mcontext __user *frame,
- struct mcontext __user *tm_frame, int ctx_has_vsx_region)
+static int __unsafe_save_user_regs(struct pt_regs *regs, struct mcontext __user *frame,
+ struct mcontext __user *tm_frame, int ctx_has_vsx_region)
{
unsigned long msr = regs->msr;
@@ -338,7 +338,7 @@ static int save_user_regs_unsafe(struct pt_regs *regs, struct mcontext __user *f
}
#define unsafe_save_user_regs(regs, frame, tm_frame, has_vsx, label) do { \
- if (save_user_regs_unsafe(regs, frame, tm_frame, has_vsx)) \
+ if (__unsafe_save_user_regs(regs, frame, tm_frame, has_vsx)) \
goto label; \
} while (0)
@@ -350,7 +350,7 @@ static int save_user_regs_unsafe(struct pt_regs *regs, struct mcontext __user *f
* We also save the transactional registers to a second ucontext in the
* frame.
*
- * See save_user_regs_unsafe() and signal_64.c:setup_tm_sigcontexts().
+ * See __unsafe_save_user_regs() and signal_64.c:setup_tm_sigcontexts().
*/
static void prepare_save_tm_user_regs(void)
{
@@ -441,7 +441,7 @@ static int save_tm_user_regs_unsafe(struct pt_regs *regs, struct mcontext __user
#endif /* CONFIG_VSX */
#ifdef CONFIG_SPE
/* SPE regs are not checkpointed with TM, so this section is
- * simply the same as in save_user_regs_unsafe().
+ * simply the same as in __unsafe_save_user_regs().
*/
if (current->thread.used_spe) {
unsafe_copy_to_user(&frame->mc_vregs, current->thread.evr,
--
2.25.0
^ permalink raw reply related
* [PATCH 06/10] powerpc/signal32: Perform access_ok() inside restore_user_regs()
From: Christophe Leroy @ 2021-03-19 11:06 UTC (permalink / raw)
To: Benjamin Herrenschmidt, Paul Mackerras, Michael Ellerman, cmr
Cc: linuxppc-dev, linux-kernel
In-Reply-To: <cover.1616151715.git.christophe.leroy@csgroup.eu>
In preparation of using user_access_begin/end in restore_user_regs(),
move the access_ok() inside the function.
It makes no difference as the behaviour on a failed access_ok() is
the same as on failed restore_user_regs().
Signed-off-by: Christophe Leroy <christophe.leroy@csgroup.eu>
---
arch/powerpc/kernel/signal_32.c | 8 +++-----
1 file changed, 3 insertions(+), 5 deletions(-)
diff --git a/arch/powerpc/kernel/signal_32.c b/arch/powerpc/kernel/signal_32.c
index 8dfe4fe77706..e2b1d2a0abad 100644
--- a/arch/powerpc/kernel/signal_32.c
+++ b/arch/powerpc/kernel/signal_32.c
@@ -492,6 +492,8 @@ static long restore_user_regs(struct pt_regs *regs,
int i;
#endif
+ if (!access_ok(sr, sizeof(*sr)))
+ return 1;
/*
* restore general registers but not including MSR or SOFTE. Also
* take care of keeping r2 (TLS) intact if not a signal
@@ -963,13 +965,10 @@ static int do_setcontext(struct ucontext __user *ucp, struct pt_regs *regs, int
if (__get_user(cmcp, &ucp->uc_regs))
return -EFAULT;
mcp = (struct mcontext __user *)(u64)cmcp;
- /* no need to check access_ok(mcp), since mcp < 4GB */
}
#else
if (__get_user(mcp, &ucp->uc_regs))
return -EFAULT;
- if (!access_ok(mcp, sizeof(*mcp)))
- return -EFAULT;
#endif
set_current_blocked(&set);
if (restore_user_regs(regs, mcp, sig))
@@ -1362,8 +1361,7 @@ SYSCALL_DEFINE0(sigreturn)
} else {
sr = (struct mcontext __user *)from_user_ptr(sigctx.regs);
addr = sr;
- if (!access_ok(sr, sizeof(*sr))
- || restore_user_regs(regs, sr, 1))
+ if (restore_user_regs(regs, sr, 1))
goto badframe;
}
--
2.25.0
^ permalink raw reply related
* [PATCH 05/10] powerpc/signal32: Remove ifdefery in middle of if/else in sigreturn()
From: Christophe Leroy @ 2021-03-19 11:06 UTC (permalink / raw)
To: Benjamin Herrenschmidt, Paul Mackerras, Michael Ellerman, cmr
Cc: linuxppc-dev, linux-kernel
In-Reply-To: <cover.1616151715.git.christophe.leroy@csgroup.eu>
In the same spirit as commit f1cf4f93de2f ("powerpc/signal32: Remove
ifdefery in middle of if/else")
MSR_TM_ACTIVE() is always defined and returns always 0 when
CONFIG_PPC_TRANSACTIONAL_MEM is not selected, so the awful
ifdefery in the middle of an if/else can be removed.
Make 'msr_hi' a 'long long' to avoid build failure on PPC32
due to the 32 bits left shift.
Signed-off-by: Christophe Leroy <christophe.leroy@csgroup.eu>
---
arch/powerpc/kernel/signal_32.c | 20 ++++++++++++--------
1 file changed, 12 insertions(+), 8 deletions(-)
diff --git a/arch/powerpc/kernel/signal_32.c b/arch/powerpc/kernel/signal_32.c
index 3b78748d6d85..8dfe4fe77706 100644
--- a/arch/powerpc/kernel/signal_32.c
+++ b/arch/powerpc/kernel/signal_32.c
@@ -740,6 +740,12 @@ static long restore_tm_user_regs(struct pt_regs *regs,
return 0;
}
+#else
+static long restore_tm_user_regs(struct pt_regs *regs, struct mcontext __user *sr,
+ struct mcontext __user *tm_sr)
+{
+ return 0;
+}
#endif
#ifdef CONFIG_PPC64
@@ -1317,10 +1323,9 @@ SYSCALL_DEFINE0(sigreturn)
struct mcontext __user *sr;
void __user *addr;
sigset_t set;
-#ifdef CONFIG_PPC_TRANSACTIONAL_MEM
- struct mcontext __user *mcp, *tm_mcp;
- unsigned long msr_hi;
-#endif
+ struct mcontext __user *mcp;
+ struct mcontext __user *tm_mcp = NULL;
+ unsigned long long msr_hi = 0;
/* Always make any pending restarted system calls return -EINTR */
current->restart_block.fn = do_no_restart_syscall;
@@ -1343,19 +1348,18 @@ SYSCALL_DEFINE0(sigreturn)
#endif
set_current_blocked(&set);
-#ifdef CONFIG_PPC_TRANSACTIONAL_MEM
mcp = (struct mcontext __user *)&sf->mctx;
+#ifdef CONFIG_PPC_TRANSACTIONAL_MEM
tm_mcp = (struct mcontext __user *)&sf->mctx_transact;
if (__get_user(msr_hi, &tm_mcp->mc_gregs[PT_MSR]))
goto badframe;
+#endif
if (MSR_TM_ACTIVE(msr_hi<<32)) {
if (!cpu_has_feature(CPU_FTR_TM))
goto badframe;
if (restore_tm_user_regs(regs, mcp, tm_mcp))
goto badframe;
- } else
-#endif
- {
+ } else {
sr = (struct mcontext __user *)from_user_ptr(sigctx.regs);
addr = sr;
if (!access_ok(sr, sizeof(*sr))
--
2.25.0
^ permalink raw reply related
* [PATCH 07/10] powerpc/signal32: Reorder user reads in restore_tm_user_regs()
From: Christophe Leroy @ 2021-03-19 11:06 UTC (permalink / raw)
To: Benjamin Herrenschmidt, Paul Mackerras, Michael Ellerman, cmr
Cc: linuxppc-dev, linux-kernel
In-Reply-To: <cover.1616151715.git.christophe.leroy@csgroup.eu>
In restore_tm_user_regs(), regroup the reads from 'sr' and the ones
from 'tm_sr' together in order to allow two block user accesses
in following patch.
Signed-off-by: Christophe Leroy <christophe.leroy@csgroup.eu>
---
arch/powerpc/kernel/signal_32.c | 49 +++++++++++++++++++++++++--------
1 file changed, 37 insertions(+), 12 deletions(-)
diff --git a/arch/powerpc/kernel/signal_32.c b/arch/powerpc/kernel/signal_32.c
index e2b1d2a0abad..088c83853026 100644
--- a/arch/powerpc/kernel/signal_32.c
+++ b/arch/powerpc/kernel/signal_32.c
@@ -607,8 +607,7 @@ static long restore_tm_user_regs(struct pt_regs *regs,
* TFHAR is restored from the checkpointed NIP; TEXASR and TFIAR
* were set by the signal delivery.
*/
- err = restore_general_regs(regs, tm_sr);
- err |= restore_general_regs(¤t->thread.ckpt_regs, sr);
+ err = restore_general_regs(¤t->thread.ckpt_regs, sr);
err |= __get_user(current->thread.tm_tfhar, &sr->mc_gregs[PT_NIP]);
@@ -624,9 +623,6 @@ static long restore_tm_user_regs(struct pt_regs *regs,
if (msr & MSR_VEC) {
/* restore altivec registers from the stack */
if (__copy_from_user(¤t->thread.ckvr_state, &sr->mc_vregs,
- sizeof(sr->mc_vregs)) ||
- __copy_from_user(¤t->thread.vr_state,
- &tm_sr->mc_vregs,
sizeof(sr->mc_vregs)))
return 1;
current->thread.used_vr = true;
@@ -639,9 +635,7 @@ static long restore_tm_user_regs(struct pt_regs *regs,
/* Always get VRSAVE back */
if (__get_user(current->thread.ckvrsave,
- (u32 __user *)&sr->mc_vregs[32]) ||
- __get_user(current->thread.vrsave,
- (u32 __user *)&tm_sr->mc_vregs[32]))
+ (u32 __user *)&sr->mc_vregs[32]))
return 1;
if (cpu_has_feature(CPU_FTR_ALTIVEC))
mtspr(SPRN_VRSAVE, current->thread.ckvrsave);
@@ -649,8 +643,7 @@ static long restore_tm_user_regs(struct pt_regs *regs,
regs->msr &= ~(MSR_FP | MSR_FE0 | MSR_FE1);
- if (copy_fpr_from_user(current, &sr->mc_fregs) ||
- copy_ckfpr_from_user(current, &tm_sr->mc_fregs))
+ if (copy_fpr_from_user(current, &sr->mc_fregs))
return 1;
#ifdef CONFIG_VSX
@@ -660,8 +653,7 @@ static long restore_tm_user_regs(struct pt_regs *regs,
* Restore altivec registers from the stack to a local
* buffer, then write this out to the thread_struct
*/
- if (copy_vsx_from_user(current, &tm_sr->mc_vsregs) ||
- copy_ckvsx_from_user(current, &sr->mc_vsregs))
+ if (copy_ckvsx_from_user(current, &sr->mc_vsregs))
return 1;
current->thread.used_vsr = true;
} else if (current->thread.used_vsr)
@@ -690,6 +682,39 @@ static long restore_tm_user_regs(struct pt_regs *regs,
return 1;
#endif /* CONFIG_SPE */
+ err = restore_general_regs(regs, tm_sr);
+ if (err)
+ return 1;
+
+#ifdef CONFIG_ALTIVEC
+ /* restore altivec registers from the stack */
+ if (msr & MSR_VEC)
+ if (__copy_from_user(¤t->thread.vr_state,
+ &tm_sr->mc_vregs,
+ sizeof(sr->mc_vregs)))
+ return 1;
+
+ /* Always get VRSAVE back */
+ if (__get_user(current->thread.vrsave,
+ (u32 __user *)&tm_sr->mc_vregs[32]))
+ return 1;
+#endif /* CONFIG_ALTIVEC */
+
+ if (copy_ckfpr_from_user(current, &tm_sr->mc_fregs))
+ return 1;
+
+#ifdef CONFIG_VSX
+ if (msr & MSR_VSX) {
+ /*
+ * Restore altivec registers from the stack to a local
+ * buffer, then write this out to the thread_struct
+ */
+ if (copy_vsx_from_user(current, &tm_sr->mc_vsregs))
+ return 1;
+ current->thread.used_vsr = true;
+ }
+#endif /* CONFIG_VSX */
+
/* Get the top half of the MSR from the user context */
if (__get_user(msr_hi, &tm_sr->mc_gregs[PT_MSR]))
return 1;
--
2.25.0
^ permalink raw reply related
* [PATCH 08/10] powerpc/signal32: Convert restore_[tm]_user_regs() to user access block
From: Christophe Leroy @ 2021-03-19 11:06 UTC (permalink / raw)
To: Benjamin Herrenschmidt, Paul Mackerras, Michael Ellerman, cmr
Cc: linuxppc-dev, linux-kernel
In-Reply-To: <cover.1616151715.git.christophe.leroy@csgroup.eu>
Convert restore_user_regs() and restore_tm_user_regs()
to use user_access_read_begin/end blocks.
Signed-off-by: Christophe Leroy <christophe.leroy@csgroup.eu>
---
arch/powerpc/include/asm/ptrace.h | 2 +-
arch/powerpc/kernel/signal_32.c | 141 +++++++++++++++---------------
2 files changed, 72 insertions(+), 71 deletions(-)
diff --git a/arch/powerpc/include/asm/ptrace.h b/arch/powerpc/include/asm/ptrace.h
index f10498e1b3f6..95600f3a6523 100644
--- a/arch/powerpc/include/asm/ptrace.h
+++ b/arch/powerpc/include/asm/ptrace.h
@@ -245,7 +245,7 @@ static inline bool trap_norestart(struct pt_regs *regs)
return regs->trap & 0x10;
}
-static inline void set_trap_norestart(struct pt_regs *regs)
+static __always_inline void set_trap_norestart(struct pt_regs *regs)
{
regs->trap |= 0x10;
}
diff --git a/arch/powerpc/kernel/signal_32.c b/arch/powerpc/kernel/signal_32.c
index 088c83853026..0b1a6f53e553 100644
--- a/arch/powerpc/kernel/signal_32.c
+++ b/arch/powerpc/kernel/signal_32.c
@@ -116,8 +116,8 @@ __unsafe_save_general_regs(struct pt_regs *regs, struct mcontext __user *frame)
return 1;
}
-static inline int restore_general_regs(struct pt_regs *regs,
- struct mcontext __user *sr)
+static __always_inline int
+__unsafe_restore_general_regs(struct pt_regs *regs, struct mcontext __user *sr)
{
elf_greg_t64 *gregs = (elf_greg_t64 *)regs;
int i;
@@ -125,10 +125,12 @@ static inline int restore_general_regs(struct pt_regs *regs,
for (i = 0; i <= PT_RESULT; i++) {
if ((i == PT_MSR) || (i == PT_SOFTE))
continue;
- if (__get_user(gregs[i], &sr->mc_gregs[i]))
- return -EFAULT;
+ unsafe_get_user(gregs[i], &sr->mc_gregs[i], failed);
}
return 0;
+
+failed:
+ return 1;
}
#else /* CONFIG_PPC64 */
@@ -161,18 +163,20 @@ __unsafe_save_general_regs(struct pt_regs *regs, struct mcontext __user *frame)
return 1;
}
-static inline int restore_general_regs(struct pt_regs *regs,
- struct mcontext __user *sr)
+static __always_inline
+int __unsafe_restore_general_regs(struct pt_regs *regs, struct mcontext __user *sr)
{
/* copy up to but not including MSR */
- if (__copy_from_user(regs, &sr->mc_gregs,
- PT_MSR * sizeof(elf_greg_t)))
- return -EFAULT;
+ unsafe_copy_from_user(regs, &sr->mc_gregs, PT_MSR * sizeof(elf_greg_t), failed);
+
/* copy from orig_r3 (the word after the MSR) up to the end */
- if (__copy_from_user(®s->orig_gpr3, &sr->mc_gregs[PT_ORIG_R3],
- GP_REGS_SIZE - PT_ORIG_R3 * sizeof(elf_greg_t)))
- return -EFAULT;
+ unsafe_copy_from_user(®s->orig_gpr3, &sr->mc_gregs[PT_ORIG_R3],
+ GP_REGS_SIZE - PT_ORIG_R3 * sizeof(elf_greg_t), failed);
+
return 0;
+
+failed:
+ return 1;
}
#endif
@@ -181,6 +185,11 @@ static inline int restore_general_regs(struct pt_regs *regs,
goto label; \
} while (0)
+#define unsafe_restore_general_regs(regs, frame, label) do { \
+ if (__unsafe_restore_general_regs(regs, frame)) \
+ goto label; \
+} while (0)
+
/*
* When we have signals to deliver, we set up on the
* user stack, going down from the original stack pointer:
@@ -485,14 +494,13 @@ static int save_tm_user_regs_unsafe(struct pt_regs *regs, struct mcontext __user
static long restore_user_regs(struct pt_regs *regs,
struct mcontext __user *sr, int sig)
{
- long err;
unsigned int save_r2 = 0;
unsigned long msr;
#ifdef CONFIG_VSX
int i;
#endif
- if (!access_ok(sr, sizeof(*sr)))
+ if (!user_read_access_begin(sr, sizeof(*sr)))
return 1;
/*
* restore general registers but not including MSR or SOFTE. Also
@@ -500,13 +508,11 @@ static long restore_user_regs(struct pt_regs *regs,
*/
if (!sig)
save_r2 = (unsigned int)regs->gpr[2];
- err = restore_general_regs(regs, sr);
+ unsafe_restore_general_regs(regs, sr, failed);
set_trap_norestart(regs);
- err |= __get_user(msr, &sr->mc_gregs[PT_MSR]);
+ unsafe_get_user(msr, &sr->mc_gregs[PT_MSR], failed);
if (!sig)
regs->gpr[2] = (unsigned long) save_r2;
- if (err)
- return 1;
/* if doing signal return, restore the previous little-endian mode */
if (sig)
@@ -520,22 +526,19 @@ static long restore_user_regs(struct pt_regs *regs,
regs->msr &= ~MSR_VEC;
if (msr & MSR_VEC) {
/* restore altivec registers from the stack */
- if (__copy_from_user(¤t->thread.vr_state, &sr->mc_vregs,
- sizeof(sr->mc_vregs)))
- return 1;
+ unsafe_copy_from_user(¤t->thread.vr_state, &sr->mc_vregs,
+ sizeof(sr->mc_vregs), failed);
current->thread.used_vr = true;
} else if (current->thread.used_vr)
memset(¤t->thread.vr_state, 0,
ELF_NVRREG * sizeof(vector128));
/* Always get VRSAVE back */
- if (__get_user(current->thread.vrsave, (u32 __user *)&sr->mc_vregs[32]))
- return 1;
+ unsafe_get_user(current->thread.vrsave, (u32 __user *)&sr->mc_vregs[32], failed);
if (cpu_has_feature(CPU_FTR_ALTIVEC))
mtspr(SPRN_VRSAVE, current->thread.vrsave);
#endif /* CONFIG_ALTIVEC */
- if (copy_fpr_from_user(current, &sr->mc_fregs))
- return 1;
+ unsafe_copy_fpr_from_user(current, &sr->mc_fregs, failed);
#ifdef CONFIG_VSX
/*
@@ -548,8 +551,7 @@ static long restore_user_regs(struct pt_regs *regs,
* Restore altivec registers from the stack to a local
* buffer, then write this out to the thread_struct
*/
- if (copy_vsx_from_user(current, &sr->mc_vsregs))
- return 1;
+ unsafe_copy_vsx_from_user(current, &sr->mc_vsregs, failed);
current->thread.used_vsr = true;
} else if (current->thread.used_vsr)
for (i = 0; i < 32 ; i++)
@@ -567,19 +569,22 @@ static long restore_user_regs(struct pt_regs *regs,
regs->msr &= ~MSR_SPE;
if (msr & MSR_SPE) {
/* restore spe registers from the stack */
- if (__copy_from_user(current->thread.evr, &sr->mc_vregs,
- ELF_NEVRREG * sizeof(u32)))
- return 1;
+ unsafe_copy_from_user(current->thread.evr, &sr->mc_vregs,
+ ELF_NEVRREG * sizeof(u32));
current->thread.used_spe = true;
} else if (current->thread.used_spe)
memset(current->thread.evr, 0, ELF_NEVRREG * sizeof(u32));
/* Always get SPEFSCR back */
- if (__get_user(current->thread.spefscr, (u32 __user *)&sr->mc_vregs + ELF_NEVRREG))
- return 1;
+ unsafe_get_user(current->thread.spefscr, (u32 __user *)&sr->mc_vregs + ELF_NEVRREG, failed);
#endif /* CONFIG_SPE */
+ user_read_access_end();
return 0;
+
+failed:
+ user_read_access_end();
+ return 1;
}
#ifdef CONFIG_PPC_TRANSACTIONAL_MEM
@@ -592,7 +597,6 @@ static long restore_tm_user_regs(struct pt_regs *regs,
struct mcontext __user *sr,
struct mcontext __user *tm_sr)
{
- long err;
unsigned long msr, msr_hi;
#ifdef CONFIG_VSX
int i;
@@ -607,14 +611,13 @@ static long restore_tm_user_regs(struct pt_regs *regs,
* TFHAR is restored from the checkpointed NIP; TEXASR and TFIAR
* were set by the signal delivery.
*/
- err = restore_general_regs(¤t->thread.ckpt_regs, sr);
-
- err |= __get_user(current->thread.tm_tfhar, &sr->mc_gregs[PT_NIP]);
-
- err |= __get_user(msr, &sr->mc_gregs[PT_MSR]);
- if (err)
+ if (!user_read_access_begin(sr, sizeof(*sr)))
return 1;
+ unsafe_restore_general_regs(¤t->thread.ckpt_regs, sr, failed);
+ unsafe_get_user(current->thread.tm_tfhar, &sr->mc_gregs[PT_NIP], failed);
+ unsafe_get_user(msr, &sr->mc_gregs[PT_MSR], failed);
+
/* Restore the previous little-endian mode */
regs->msr = (regs->msr & ~MSR_LE) | (msr & MSR_LE);
@@ -622,9 +625,8 @@ static long restore_tm_user_regs(struct pt_regs *regs,
regs->msr &= ~MSR_VEC;
if (msr & MSR_VEC) {
/* restore altivec registers from the stack */
- if (__copy_from_user(¤t->thread.ckvr_state, &sr->mc_vregs,
- sizeof(sr->mc_vregs)))
- return 1;
+ unsafe_copy_from_user(¤t->thread.ckvr_state, &sr->mc_vregs,
+ sizeof(sr->mc_vregs), failed);
current->thread.used_vr = true;
} else if (current->thread.used_vr) {
memset(¤t->thread.vr_state, 0,
@@ -634,17 +636,15 @@ static long restore_tm_user_regs(struct pt_regs *regs,
}
/* Always get VRSAVE back */
- if (__get_user(current->thread.ckvrsave,
- (u32 __user *)&sr->mc_vregs[32]))
- return 1;
+ unsafe_get_user(current->thread.ckvrsave,
+ (u32 __user *)&sr->mc_vregs[32], failed);
if (cpu_has_feature(CPU_FTR_ALTIVEC))
mtspr(SPRN_VRSAVE, current->thread.ckvrsave);
#endif /* CONFIG_ALTIVEC */
regs->msr &= ~(MSR_FP | MSR_FE0 | MSR_FE1);
- if (copy_fpr_from_user(current, &sr->mc_fregs))
- return 1;
+ unsafe_copy_fpr_from_user(current, &sr->mc_fregs, failed);
#ifdef CONFIG_VSX
regs->msr &= ~MSR_VSX;
@@ -653,8 +653,7 @@ static long restore_tm_user_regs(struct pt_regs *regs,
* Restore altivec registers from the stack to a local
* buffer, then write this out to the thread_struct
*/
- if (copy_ckvsx_from_user(current, &sr->mc_vsregs))
- return 1;
+ unsafe_copy_ckvsx_from_user(current, &sr->mc_vsregs, failed);
current->thread.used_vsr = true;
} else if (current->thread.used_vsr)
for (i = 0; i < 32 ; i++) {
@@ -669,39 +668,36 @@ static long restore_tm_user_regs(struct pt_regs *regs,
*/
regs->msr &= ~MSR_SPE;
if (msr & MSR_SPE) {
- if (__copy_from_user(current->thread.evr, &sr->mc_vregs,
- ELF_NEVRREG * sizeof(u32)))
- return 1;
+ unsafe_copy_from_user(current->thread.evr, &sr->mc_vregs,
+ ELF_NEVRREG * sizeof(u32), failed);
current->thread.used_spe = true;
} else if (current->thread.used_spe)
memset(current->thread.evr, 0, ELF_NEVRREG * sizeof(u32));
/* Always get SPEFSCR back */
- if (__get_user(current->thread.spefscr, (u32 __user *)&sr->mc_vregs
- + ELF_NEVRREG))
- return 1;
+ unsafe_get_user(current->thread.spefscr,
+ (u32 __user *)&sr->mc_vregs + ELF_NEVRREG, failed);
#endif /* CONFIG_SPE */
- err = restore_general_regs(regs, tm_sr);
- if (err)
+ user_read_access_end();
+
+ if (!user_read_access_begin(tm_sr, sizeof(*tm_sr)))
return 1;
+ unsafe_restore_general_regs(regs, tm_sr, failed);
+
#ifdef CONFIG_ALTIVEC
/* restore altivec registers from the stack */
if (msr & MSR_VEC)
- if (__copy_from_user(¤t->thread.vr_state,
- &tm_sr->mc_vregs,
- sizeof(sr->mc_vregs)))
- return 1;
+ unsafe_copy_from_user(¤t->thread.vr_state, &tm_sr->mc_vregs,
+ sizeof(sr->mc_vregs), failed);
/* Always get VRSAVE back */
- if (__get_user(current->thread.vrsave,
- (u32 __user *)&tm_sr->mc_vregs[32]))
- return 1;
+ unsafe_get_user(current->thread.vrsave,
+ (u32 __user *)&tm_sr->mc_vregs[32], failed);
#endif /* CONFIG_ALTIVEC */
- if (copy_ckfpr_from_user(current, &tm_sr->mc_fregs))
- return 1;
+ unsafe_copy_ckfpr_from_user(current, &tm_sr->mc_fregs, failed);
#ifdef CONFIG_VSX
if (msr & MSR_VSX) {
@@ -709,16 +705,17 @@ static long restore_tm_user_regs(struct pt_regs *regs,
* Restore altivec registers from the stack to a local
* buffer, then write this out to the thread_struct
*/
- if (copy_vsx_from_user(current, &tm_sr->mc_vsregs))
- return 1;
+ unsafe_copy_vsx_from_user(current, &tm_sr->mc_vsregs, failed);
current->thread.used_vsr = true;
}
#endif /* CONFIG_VSX */
/* Get the top half of the MSR from the user context */
- if (__get_user(msr_hi, &tm_sr->mc_gregs[PT_MSR]))
- return 1;
+ unsafe_get_user(msr_hi, &tm_sr->mc_gregs[PT_MSR], failed);
msr_hi <<= 32;
+
+ user_read_access_end();
+
/* If TM bits are set to the reserved value, it's an invalid context */
if (MSR_TM_RESV(msr_hi))
return 1;
@@ -766,6 +763,10 @@ static long restore_tm_user_regs(struct pt_regs *regs,
preempt_enable();
return 0;
+
+failed:
+ user_read_access_end();
+ return 1;
}
#else
static long restore_tm_user_regs(struct pt_regs *regs, struct mcontext __user *sr,
--
2.25.0
^ permalink raw reply related
* [PATCH 10/10] powerpc/signal32: Simplify logging in sigreturn()
From: Christophe Leroy @ 2021-03-19 11:07 UTC (permalink / raw)
To: Benjamin Herrenschmidt, Paul Mackerras, Michael Ellerman, cmr
Cc: linuxppc-dev, linux-kernel
In-Reply-To: <cover.1616151715.git.christophe.leroy@csgroup.eu>
Same spirit as commit debf122c777f ("powerpc/signal32: Simplify logging
in handle_rt_signal32()"), remove this intermediate 'addr' local var.
Signed-off-by: Christophe Leroy <christophe.leroy@csgroup.eu>
---
arch/powerpc/kernel/signal_32.c | 13 +++++++------
1 file changed, 7 insertions(+), 6 deletions(-)
diff --git a/arch/powerpc/kernel/signal_32.c b/arch/powerpc/kernel/signal_32.c
index 592b889e3836..5be267b3a13e 100644
--- a/arch/powerpc/kernel/signal_32.c
+++ b/arch/powerpc/kernel/signal_32.c
@@ -1352,7 +1352,6 @@ SYSCALL_DEFINE0(sigreturn)
struct sigcontext __user *sc;
struct sigcontext sigctx;
struct mcontext __user *sr;
- void __user *addr;
sigset_t set;
struct mcontext __user *mcp;
struct mcontext __user *tm_mcp = NULL;
@@ -1363,7 +1362,6 @@ SYSCALL_DEFINE0(sigreturn)
sf = (struct sigframe __user *)(regs->gpr[1] + __SIGNAL_FRAMESIZE);
sc = &sf->sctx;
- addr = sc;
if (copy_from_user(&sigctx, sc, sizeof(sigctx)))
goto badframe;
@@ -1392,16 +1390,19 @@ SYSCALL_DEFINE0(sigreturn)
goto badframe;
} else {
sr = (struct mcontext __user *)from_user_ptr(sigctx.regs);
- addr = sr;
- if (restore_user_regs(regs, sr, 1))
- goto badframe;
+ if (restore_user_regs(regs, sr, 1)) {
+ signal_fault(current, regs, "sys_sigreturn", sr);
+
+ force_sig(SIGSEGV);
+ return 0;
+ }
}
set_thread_flag(TIF_RESTOREALL);
return 0;
badframe:
- signal_fault(current, regs, "sys_sigreturn", addr);
+ signal_fault(current, regs, "sys_sigreturn", sc);
force_sig(SIGSEGV);
return 0;
--
2.25.0
^ permalink raw reply related
* [PATCH 09/10] powerpc/signal32: Convert do_setcontext[_tm]() to user access block
From: Christophe Leroy @ 2021-03-19 11:06 UTC (permalink / raw)
To: Benjamin Herrenschmidt, Paul Mackerras, Michael Ellerman, cmr
Cc: linuxppc-dev, linux-kernel
In-Reply-To: <cover.1616151715.git.christophe.leroy@csgroup.eu>
Add unsafe_get_user_sigset() and transform PPC32 get_sigset_t()
into an unsafe version unsafe_get_sigset_t().
Then convert do_setcontext() and do_setcontext_tm() to use
user_read_access_begin/end.
Signed-off-by: Christophe Leroy <christophe.leroy@csgroup.eu>
---
arch/powerpc/kernel/signal.h | 2 ++
arch/powerpc/kernel/signal_32.c | 42 +++++++++++++++++++--------------
2 files changed, 26 insertions(+), 18 deletions(-)
diff --git a/arch/powerpc/kernel/signal.h b/arch/powerpc/kernel/signal.h
index a5152ff3c52f..f4aafa337c2e 100644
--- a/arch/powerpc/kernel/signal.h
+++ b/arch/powerpc/kernel/signal.h
@@ -25,6 +25,8 @@ static inline int __get_user_sigset(sigset_t *dst, const sigset_t __user *src)
return __get_user(dst->sig[0], (u64 __user *)&src->sig[0]);
}
+#define unsafe_get_user_sigset(dst, src, label) \
+ unsafe_get_user((dst)->sig[0], (u64 __user *)&(src)->sig[0], label)
#ifdef CONFIG_VSX
extern unsigned long copy_vsx_to_user(void __user *to,
diff --git a/arch/powerpc/kernel/signal_32.c b/arch/powerpc/kernel/signal_32.c
index 0b1a6f53e553..592b889e3836 100644
--- a/arch/powerpc/kernel/signal_32.c
+++ b/arch/powerpc/kernel/signal_32.c
@@ -83,12 +83,7 @@
* implementation that makes things simple for little endian only)
*/
#define unsafe_put_sigset_t unsafe_put_compat_sigset
-
-static inline int get_sigset_t(sigset_t *set,
- const compat_sigset_t __user *uset)
-{
- return get_compat_sigset(set, uset);
-}
+#define unsafe_get_sigset_t unsafe_get_compat_sigset
#define to_user_ptr(p) ptr_to_compat(p)
#define from_user_ptr(p) compat_ptr(p)
@@ -144,10 +139,7 @@ __unsafe_restore_general_regs(struct pt_regs *regs, struct mcontext __user *sr)
unsafe_copy_to_user(__us, __s, sizeof(*__us), label); \
} while (0)
-static inline int get_sigset_t(sigset_t *set, const sigset_t __user *uset)
-{
- return __get_user_sigset(set, uset);
-}
+#define unsafe_get_sigset_t unsafe_get_user_sigset
#define to_user_ptr(p) ((unsigned long)(p))
#define from_user_ptr(p) ((void __user *)(p))
@@ -982,25 +974,31 @@ static int do_setcontext(struct ucontext __user *ucp, struct pt_regs *regs, int
sigset_t set;
struct mcontext __user *mcp;
- if (get_sigset_t(&set, &ucp->uc_sigmask))
+ if (user_read_access_begin(ucp, sizeof(*ucp)))
return -EFAULT;
+
+ unsafe_get_sigset_t(&set, &ucp->uc_sigmask, failed);
#ifdef CONFIG_PPC64
{
u32 cmcp;
- if (__get_user(cmcp, &ucp->uc_regs))
- return -EFAULT;
+ unsafe_get_user(cmcp, &ucp->uc_regs, failed);
mcp = (struct mcontext __user *)(u64)cmcp;
}
#else
- if (__get_user(mcp, &ucp->uc_regs))
- return -EFAULT;
+ unsafe_get_user(mcp, &ucp->uc_regs, failed);
#endif
+ user_read_access_end();
+
set_current_blocked(&set);
if (restore_user_regs(regs, mcp, sig))
return -EFAULT;
return 0;
+
+failed:
+ user_read_access_end();
+ return -EFAULT;
}
#ifdef CONFIG_PPC_TRANSACTIONAL_MEM
@@ -1014,11 +1012,15 @@ static int do_setcontext_tm(struct ucontext __user *ucp,
u32 cmcp;
u32 tm_cmcp;
- if (get_sigset_t(&set, &ucp->uc_sigmask))
+ if (user_read_access_begin(ucp, sizeof(*ucp)))
return -EFAULT;
- if (__get_user(cmcp, &ucp->uc_regs) ||
- __get_user(tm_cmcp, &tm_ucp->uc_regs))
+ unsafe_get_sigset_t(&set, &ucp->uc_sigmask, failed);
+ unsafe_get_user(cmcp, &ucp->uc_regs, failed);
+
+ user_read_access_end();
+
+ if (__get_user(tm_cmcp, &tm_ucp->uc_regs))
return -EFAULT;
mcp = (struct mcontext __user *)(u64)cmcp;
tm_mcp = (struct mcontext __user *)(u64)tm_cmcp;
@@ -1029,6 +1031,10 @@ static int do_setcontext_tm(struct ucontext __user *ucp,
return -EFAULT;
return 0;
+
+failed:
+ user_read_access_end();
+ return -EFAULT;
}
#endif
--
2.25.0
^ permalink raw reply related
* Re: [PATCH 1/1] hotplug-cpu.c: show 'last online CPU' error in dlpar_cpu_remove()
From: Michael Ellerman @ 2021-03-19 11:26 UTC (permalink / raw)
To: Daniel Henrique Barboza, linuxppc-dev
In-Reply-To: <c2fdbf4e-995b-fec8-7dc3-41beb7d66daa@gmail.com>
Daniel Henrique Barboza <danielhb413@gmail.com> writes:
> Ping
>
> On 3/5/21 2:38 PM, Daniel Henrique Barboza wrote:
>> Of all the reasons that dlpar_cpu_remove() can fail, the 'last online
>> CPU' is one that can be caused directly by the user offlining CPUs
>> in a partition/virtual machine that has hotplugged CPUs. Trying to
>> reclaim a hotplugged CPU can fail if the CPU is now the last online in
>> the system. This is easily reproduced using QEMU [1].
Sorry, I saw this earlier and never got around to replying.
I'm wondering if we neet to catch it earlier, ie. in
dlpar_offline_cpu().
By the time we return to dlpar_cpu_remove() we've dropped the
cpu_add_remove_lock (cpu_maps_update_done), so num_online_cpus() could
change out from under us, meaning the num_online_cpus() == 1 check might
trigger incorrectly (or vice versa).
Something like the patch below (completely untested :D)
And writing that patch makes me wonder, is == 1 the right check?
In most cases we'll remove all but one thread of the core, but we'll
fail on the last thread. Leaving that core effectively stuck in SMT1. Is
that useful behaviour? Should we instead check at the start that we can
remove all threads of the core without going to zero online CPUs?
cheers
diff --git a/arch/powerpc/platforms/pseries/hotplug-cpu.c b/arch/powerpc/platforms/pseries/hotplug-cpu.c
index 12cbffd3c2e3..498c22331ac8 100644
--- a/arch/powerpc/platforms/pseries/hotplug-cpu.c
+++ b/arch/powerpc/platforms/pseries/hotplug-cpu.c
@@ -271,6 +271,12 @@ static int dlpar_offline_cpu(struct device_node *dn)
if (!cpu_online(cpu))
break;
+ if (num_online_cpus() == 1) {
+ pr_warn("Unable to remove last online CPU %pOFn\n", dn);
+ rc = EBUSY;
+ goto out_unlock;
+ }
+
cpu_maps_update_done();
rc = device_offline(get_cpu_device(cpu));
if (rc)
@@ -283,6 +289,7 @@ static int dlpar_offline_cpu(struct device_node *dn)
thread);
}
}
+out_unlock:
cpu_maps_update_done();
out:
^ permalink raw reply related
* Re: [PATCH 12/14] powerpc/64s: system call avoid setting MSR[RI] until we set MSR[EE]
From: Michael Ellerman @ 2021-03-19 11:29 UTC (permalink / raw)
To: Christophe Leroy, Nicholas Piggin, linuxppc-dev
In-Reply-To: <722aeb8d-507f-6702-dd79-26242f987e3e@csgroup.eu>
Christophe Leroy <christophe.leroy@csgroup.eu> writes:
> Le 15/03/2021 à 23:04, Nicholas Piggin a écrit :
>> This extends the MSR[RI]=0 window a little further into the system
>> call in order to pair RI and EE enabling with a single mtmsrd.
>
> Time ago, I proposed to delay that on PPC32 and Michael objected, see
> https://patchwork.ozlabs.org/project/linuxppc-dev/patch/9f9dd859d571e324c7412ed9db9da8cfba678257.1548956511.git.christophe.leroy@c-s.fr/
I don't think I objected, I was just curious about what the added
exposure to RI=0 was :)
cheers
>> diff --git a/arch/powerpc/kernel/exceptions-64s.S b/arch/powerpc/kernel/exceptions-64s.S
>> index bd0c82ac9de5..2f14ac3c377c 100644
>> --- a/arch/powerpc/kernel/exceptions-64s.S
>> +++ b/arch/powerpc/kernel/exceptions-64s.S
>> @@ -1999,8 +1999,6 @@ END_FTR_SECTION_IFSET(CPU_FTR_REAL_LE)
>> mtctr r10
>> bctr
>> .else
>> - li r10,MSR_RI
>> - mtmsrd r10,1 /* Set RI (EE=0) */
>> #ifdef CONFIG_RELOCATABLE
>> __LOAD_HANDLER(r10, system_call_common)
>> mtctr r10
>> diff --git a/arch/powerpc/kernel/interrupt_64.S b/arch/powerpc/kernel/interrupt_64.S
>> index f28f41a1a85a..eef61800f734 100644
>> --- a/arch/powerpc/kernel/interrupt_64.S
>> +++ b/arch/powerpc/kernel/interrupt_64.S
>> @@ -311,10 +311,10 @@ END_BTB_FLUSH_SECTION
>> * nothing pending. system_call_exception() will call
>> * trace_hardirqs_off().
>> */
>> - li r11,IRQS_ALL_DISABLED
>> - li r12,PACA_IRQ_HARD_DIS
>> + li r11,IRQS_DISABLED
>> + li r12,-1 /* Set MSR_EE and MSR_RI */
>> stb r11,PACAIRQSOFTMASK(r13)
>> - stb r12,PACAIRQHAPPENED(r13)
>> + mtmsrd r12,1
>>
>> ENTER_KERNEL_SECURITY_FALLBACK
>>
>>
^ permalink raw reply
* Re: [PATCH v3 28/32] powerpc/64s: interrupt implement exit logic in C
From: Michael Ellerman @ 2021-03-19 11:44 UTC (permalink / raw)
To: Nicholas Piggin, Christophe Leroy, linuxppc-dev; +Cc: Michal Suchanek
In-Reply-To: <1615879834.64tfygznle.astroid@bobo.none>
Nicholas Piggin <npiggin@gmail.com> writes:
> Excerpts from Christophe Leroy's message of March 15, 2021 11:41 pm:
>>
>> Le 25/02/2020 à 18:35, Nicholas Piggin a écrit :
>>> Implement the bulk of interrupt return logic in C. The asm return code
>>> must handle a few cases: restoring full GPRs, and emulating stack store.
>>>
>>> The stack store emulation is significantly simplfied, rather than creating
>>> a new return frame and switching to that before performing the store, it
>>> uses the PACA to keep a scratch register around to perform thestore.
>>>
>>> The asm return code is moved into 64e for now. The new logic has made
>>> allowance for 64e, but I don't have a full environment that works well
>>> to test it, and even booting in emulated qemu is not great for stress
>>> testing. 64e shouldn't be too far off working with this, given a bit
>>> more testing and auditing of the logic.
>>>
>>> This is slightly faster on a POWER9 (page fault speed increases about
>>> 1.1%), probably due to reduced mtmsrd.
>>>
>>> Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
>>> Signed-off-by: Michal Suchanek <msuchanek@suse.de>
>>> ---
>>
>> ...
>>
>>> +notrace unsigned long interrupt_exit_user_prepare(struct pt_regs *regs, unsigned long msr)
>>> +{
>>
>> ...
>>
>>> +
>>> +#ifdef CONFIG_PPC_TRANSACTIONAL_MEM
>>> + local_paca->tm_scratch = regs->msr;
>>> +#endif
>>
>> Could we define a helper for that in asm/tm.h, that voids when CONFIG_PPC_TRANSACTIONAL_MEM is not
>> selected ?
>
> Yeah I wanted to do something about that. I don't know what it's used
> for here. I guess it saves the return MSR so if that causes a crash then
> the next oops would see it, but I wonder if we can just get that from
> SRR1 + program check error codes, or if there is something we can't
> reconstruct from there.
In the cases when you need it, you can't reconstruct it :)
But given the TM code is on life support we could probably drop
tm_scratch.
I don't think we've used it in anger for several years. Probably since
265e60a170d0 ("powerpc/64s: Use emergency stack for kernel TM Bad Thing
program checks") (Oct 2017).
If one of us has to debug some hairy TM issue we can always add it back
temporarily in a dev kernel.
cheers
^ permalink raw reply
* Re: [for-stable-4.19 PATCH 1/2] vmlinux.lds.h: Create section for protection against instrumentation
From: Greg Kroah-Hartman @ 2021-03-19 11:55 UTC (permalink / raw)
To: Alexandre Chartre
Cc: Sasha Levin, linux-arch, Nicolas Boichat, Arnd Bergmann,
linux-kbuild, Peter Zijlstra, Christopher Li, linux-kernel,
stable, Masahiro Yamada, linux-sparse, Michal Marek,
Paul Mackerras, Nicholas Piggin, Thomas Gleixner, linuxppc-dev,
Naveen N. Rao, Daniel Axtens
In-Reply-To: <b5d3d0ed-953e-083d-15f6-4a1e3ed95428@oracle.com>
On Fri, Mar 19, 2021 at 12:20:22PM +0100, Alexandre Chartre wrote:
>
> On 3/19/21 11:39 AM, Greg Kroah-Hartman wrote:
> > On Fri, Mar 19, 2021 at 07:54:15AM +0800, Nicolas Boichat wrote:
> > > From: Thomas Gleixner <tglx@linutronix.de>
> > >
> > > commit 6553896666433e7efec589838b400a2a652b3ffa upstream.
> > >
> > > Some code pathes, especially the low level entry code, must be protected
> > > against instrumentation for various reasons:
> > >
> > > - Low level entry code can be a fragile beast, especially on x86.
> > >
> > > - With NO_HZ_FULL RCU state needs to be established before using it.
> > >
> > > Having a dedicated section for such code allows to validate with tooling
> > > that no unsafe functions are invoked.
> > >
> > > Add the .noinstr.text section and the noinstr attribute to mark
> > > functions. noinstr implies notrace. Kprobes will gain a section check
> > > later.
> > >
> > > Provide also a set of markers: instrumentation_begin()/end()
> > >
> > > These are used to mark code inside a noinstr function which calls
> > > into regular instrumentable text section as safe.
> > >
> > > The instrumentation markers are only active when CONFIG_DEBUG_ENTRY is
> > > enabled as the end marker emits a NOP to prevent the compiler from merging
> > > the annotation points. This means the objtool verification requires a
> > > kernel compiled with this option.
> > >
> > > Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
> > > Reviewed-by: Alexandre Chartre <alexandre.chartre@oracle.com>
> > > Acked-by: Peter Zijlstra <peterz@infradead.org>
> > > Link: https://lkml.kernel.org/r/20200505134100.075416272@linutronix.de
> > >
> > > [Nicolas: context conflicts in:
> > > arch/powerpc/kernel/vmlinux.lds.S
> > > include/asm-generic/vmlinux.lds.h
> > > include/linux/compiler.h
> > > include/linux/compiler_types.h]
> > > Signed-off-by: Nicolas Boichat <drinkcat@chromium.org>
> >
> > Did you build this on x86?
> >
> > I get the following build error:
> >
> > ld:./arch/x86/kernel/vmlinux.lds:20: syntax error
> >
> > And that line looks like:
> >
> > . = ALIGN(8); *(.text.hot .text.hot.*) *(.text .text.fixup) *(.text.unlikely .text.unlikely.*) *(.text.unknown .text.unknown.*) . = ALIGN(8); __noinstr_text_start = .; *(.__attribute__((noinline)) __attribute__((no_instrument_function)) __attribute((__section__(".noinstr.text"))).text) __noinstr_text_end = .; *(.text..refcount) *(.ref.text) *(.meminit.text*) *(.memexit.text*)
> >
>
> In the NOINSTR_TEXT macro, noinstr is expanded with the value of the noinstr
> macro from linux/compiler_types.h while it shouldn't.
>
> The problem is possibly that the noinstr macro is defined for assembly. Make
> sure that the macro is not defined for assembly e.g.:
>
> #ifndef __ASSEMBLY__
>
> /* Section for code which can't be instrumented at all */
> #define noinstr \
> noinline notrace __attribute((__section__(".noinstr.text")))
>
> #endif
This implies that the backport is incorrect, so I'll wait for an updated
version...
thanks,
greg k-h
^ permalink raw reply
* Re: [PATCH 5/6] powerpc/mm/64s/hash: Add real-mode change_memory_range() for hash LPAR
From: Michael Ellerman @ 2021-03-19 11:56 UTC (permalink / raw)
To: Daniel Axtens, linuxppc-dev; +Cc: aneesh.kumar
In-Reply-To: <87h7m8pyk5.fsf@dja-thinkpad.axtens.net>
Daniel Axtens <dja@axtens.net> writes:
> Michael Ellerman <mpe@ellerman.id.au> writes:
>
>> When we enabled STRICT_KERNEL_RWX we received some reports of boot
>> failures when using the Hash MMU and running under phyp. The crashes
>> are intermittent, and often exhibit as a completely unresponsive
>> system, or possibly an oops.
>>
>> One example, which was caught in xmon:
>>
>> [ 14.068327][ T1] devtmpfs: mounted
>> [ 14.069302][ T1] Freeing unused kernel memory: 5568K
>> [ 14.142060][ T347] BUG: Unable to handle kernel instruction fetch
>> [ 14.142063][ T1] Run /sbin/init as init process
>> [ 14.142074][ T347] Faulting instruction address: 0xc000000000004400
>> cpu 0x2: Vector: 400 (Instruction Access) at [c00000000c7475e0]
>> pc: c000000000004400: exc_virt_0x4400_instruction_access+0x0/0x80
>> lr: c0000000001862d4: update_rq_clock+0x44/0x110
>> sp: c00000000c747880
>> msr: 8000000040001031
>> current = 0xc00000000c60d380
>> paca = 0xc00000001ec9de80 irqmask: 0x03 irq_happened: 0x01
>> pid = 347, comm = kworker/2:1
>> ...
>> enter ? for help
>> [c00000000c747880] c0000000001862d4 update_rq_clock+0x44/0x110 (unreliable)
>> [c00000000c7478f0] c000000000198794 update_blocked_averages+0xb4/0x6d0
>> [c00000000c7479f0] c000000000198e40 update_nohz_stats+0x90/0xd0
>> [c00000000c747a20] c0000000001a13b4 _nohz_idle_balance+0x164/0x390
>> [c00000000c747b10] c0000000001a1af8 newidle_balance+0x478/0x610
>> [c00000000c747be0] c0000000001a1d48 pick_next_task_fair+0x58/0x480
>> [c00000000c747c40] c000000000eaab5c __schedule+0x12c/0x950
>> [c00000000c747cd0] c000000000eab3e8 schedule+0x68/0x120
>> [c00000000c747d00] c00000000016b730 worker_thread+0x130/0x640
>> [c00000000c747da0] c000000000174d50 kthread+0x1a0/0x1b0
>> [c00000000c747e10] c00000000000e0f0 ret_from_kernel_thread+0x5c/0x6c
>>
>> This shows that CPU 2, which was idle, woke up and then appears to
>> randomly take an instruction fault on a completely valid area of
>> kernel text.
>>
>> The cause turns out to be the call to hash__mark_rodata_ro(), late in
>> boot. Due to the way we layout text and rodata, that function actually
>> changes the permissions for all of text and rodata to read-only plus
>> execute.
>>
>> To do the permission change we use a hypervisor call, H_PROTECT. On
>> phyp that appears to be implemented by briefly removing the mapping of
>> the kernel text, before putting it back with the updated permissions.
>> If any other CPU is executing during that window, it will see spurious
>> faults on the kernel text and/or data, leading to crashes.
>
> Jordan asked why we saw this on phyp but not under KVM? We had a look at
> book3s_hv_rm_mmu.c but the code is a bit too obtuse for me to reason
> about!
>
> Nick suggests that the KVM hypervisor is invalidating the HPTE, but
> because we run guests in VPM mode, the hypervisor would catch the page
> fault and not reflect it down to the guest. It looks like Linux-as-a-HV
> will take HPTE_V_HVLOCK, and then because it's running in VPM mode, the
> hypervisor will catch the fault and not pass it to the guest.
Yep.
> But if phyp runs with VPM mode off, the guest will see the fault
> before the hypervisor. (we think this is what's going on anyway.)
Yeah. I assumed phyp always ran with VPM=1, but apparently it can run
with it off or on, depending on various configuration settings.
So I'm fairly sure what we're hitting here is VPM=0, where the faults go
straight to the guest.
cheers
^ permalink raw reply
* Re: [PATCH V2] mm/memtest: Add ARCH_USE_MEMTEST
From: Catalin Marinas @ 2021-03-19 12:34 UTC (permalink / raw)
To: Anshuman Khandual
Cc: Chris Zankel, Thomas Bogendoerfer, linux-kernel, Max Filippov,
linux-xtensa, linuxppc-dev, Russell King, linux-mips, linux-mm,
Ingo Molnar, Paul Mackerras, Thomas Gleixner, Will Deacon,
linux-arm-kernel
In-Reply-To: <1614573126-7740-1-git-send-email-anshuman.khandual@arm.com>
On Mon, Mar 01, 2021 at 10:02:06AM +0530, Anshuman Khandual wrote:
> early_memtest() does not get called from all architectures. Hence enabling
> CONFIG_MEMTEST and providing a valid memtest=[1..N] kernel command line
> option might not trigger the memory pattern tests as would be expected in
> normal circumstances. This situation is misleading.
>
> The change here prevents the above mentioned problem after introducing a
> new config option ARCH_USE_MEMTEST that should be subscribed on platforms
> that call early_memtest(), in order to enable the config CONFIG_MEMTEST.
> Conversely CONFIG_MEMTEST cannot be enabled on platforms where it would
> not be tested anyway.
>
> Cc: Russell King <linux@armlinux.org.uk>
> Cc: Catalin Marinas <catalin.marinas@arm.com>
> Cc: Will Deacon <will@kernel.org>
> Cc: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
> Cc: Michael Ellerman <mpe@ellerman.id.au>
> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
> Cc: Paul Mackerras <paulus@samba.org>
> Cc: Thomas Gleixner <tglx@linutronix.de>
> Cc: Ingo Molnar <mingo@redhat.com>
> Cc: Chris Zankel <chris@zankel.net>
> Cc: Max Filippov <jcmvbkbc@gmail.com>
> Cc: linux-arm-kernel@lists.infradead.org
> Cc: linux-mips@vger.kernel.org
> Cc: linuxppc-dev@lists.ozlabs.org
> Cc: linux-xtensa@linux-xtensa.org
> Cc: linux-mm@kvack.org
> Cc: linux-kernel@vger.kernel.org
> Reviewed-by: Max Filippov <jcmvbkbc@gmail.com>
> Signed-off-by: Anshuman Khandual <anshuman.khandual@arm.com>
For arm64:
Acked-by: Catalin Marinas <catalin.marinas@arm.com>
^ permalink raw reply
* Re: [PATCH 08/10] MIPS: disable CONFIG_IDE in malta*_defconfig
From: Christoph Hellwig @ 2021-03-19 13:03 UTC (permalink / raw)
To: Thomas Bogendoerfer
Cc: Jens Axboe, linux-doc, Russell King, David S. Miller, linux-ide,
linux-m68k, Ivan Kokshaysky, linux-arm-kernel, linux-alpha,
Geert Uytterhoeven, Matt Turner, linux-mips, linuxppc-dev,
Christoph Hellwig, linux-kernel, Richard Henderson
In-Reply-To: <20210318141900.GA10554@alpha.franken.de>
On Thu, Mar 18, 2021 at 03:19:00PM +0100, Thomas Bogendoerfer wrote:
> On Thu, Mar 18, 2021 at 05:57:04AM +0100, Christoph Hellwig wrote:
> > arch/mips/configs/malta_kvm_guest_defconfig | 3 ---
>
> that file is gone in mips-next.
>
> I could take all MIPS patches into mips-next, if you want...
Fine with me - it shouldn't really matter if the defconfig updates
go in independently. Do you want a resend with the typos fixed
against mips-next or are you simply going to fix these things up?
^ permalink raw reply
* [PATCH RFC 2/3] mm: remove xlate_dev_kmem_ptr()
From: David Hildenbrand @ 2021-03-19 14:34 UTC (permalink / raw)
To: linux-kernel
Cc: Rich Felker, linux-ia64, Kuninori Morimoto, David Hildenbrand,
Peter Zijlstra (Intel), Sebastian Andrzej Siewior, Palmer Dabbelt,
Jiaxun Yang, James E.J. Bottomley, linux-mm, Paul Mackerras,
sparclinux, linux-hexagon, Ingo Molnar, linux-arch, linux-s390,
Vasily Gorbik, Brian Cain, Helge Deller, linux-sh, Russell King,
Krzysztof Kozlowski, Christian Borntraeger, Geert Uytterhoeven,
linux-parisc, Matt Turner, Arnd Bergmann, Niklas Schnelle,
Heiko Carstens, linux-m68k, Ivan Kokshaysky, Greentime Hu,
Gerald Schaefer, linux-arm-kernel, Richard Henderson,
Thomas Bogendoerfer, Yoshinori Sato, Pierre Morel, Randy Dunlap,
linux-mips, Luis Chamberlain, Mike Rapoport, linux-alpha,
Mikulas Patocka, Andrew Morton, linuxppc-dev, David S. Miller,
Luc Van Oostenryck
In-Reply-To: <20210319143452.25948-1-david@redhat.com>
Since /dev/kmem has been removed, let's remove the xlate_dev_kmem_ptr()
leftovers.
Cc: Richard Henderson <rth@twiddle.net>
Cc: Ivan Kokshaysky <ink@jurassic.park.msu.ru>
Cc: Matt Turner <mattst88@gmail.com>
Cc: Russell King <linux@armlinux.org.uk>
Cc: Brian Cain <bcain@codeaurora.org>
Cc: Geert Uytterhoeven <geert@linux-m68k.org>
Cc: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Cc: "James E.J. Bottomley" <James.Bottomley@HansenPartnership.com>
Cc: Helge Deller <deller@gmx.de>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Heiko Carstens <hca@linux.ibm.com>
Cc: Vasily Gorbik <gor@linux.ibm.com>
Cc: Christian Borntraeger <borntraeger@de.ibm.com>
Cc: Yoshinori Sato <ysato@users.sourceforge.jp>
Cc: Rich Felker <dalias@libc.org>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: David Hildenbrand <david@redhat.com>
Cc: Krzysztof Kozlowski <krzk@kernel.org>
Cc: Mikulas Patocka <mpatocka@redhat.com>
Cc: Luc Van Oostenryck <luc.vanoostenryck@gmail.com>
Cc: Mike Rapoport <rppt@kernel.org>
Cc: Palmer Dabbelt <palmerdabbelt@google.com>
Cc: Luis Chamberlain <mcgrof@kernel.org>
Cc: Greentime Hu <green.hu@gmail.com>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: Randy Dunlap <rdunlap@infradead.org>
Cc: Jiaxun Yang <jiaxun.yang@flygoat.com>
Cc: "Peter Zijlstra (Intel)" <peterz@infradead.org>
Cc: Christophe Leroy <christophe.leroy@csgroup.eu>
Cc: Gerald Schaefer <gerald.schaefer@linux.ibm.com>
Cc: Niklas Schnelle <schnelle@linux.ibm.com>
Cc: Pierre Morel <pmorel@linux.ibm.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Cc: linux-alpha@vger.kernel.org
Cc: linux-arm-kernel@lists.infradead.org
Cc: linux-hexagon@vger.kernel.org
Cc: linux-ia64@vger.kernel.org
Cc: linux-m68k@lists.linux-m68k.org
Cc: linux-mips@vger.kernel.org
Cc: linux-parisc@vger.kernel.org
Cc: linuxppc-dev@lists.ozlabs.org
Cc: linux-s390@vger.kernel.org
Cc: linux-sh@vger.kernel.org
Cc: sparclinux@vger.kernel.org
Cc: linux-arch@vger.kernel.org
Signed-off-by: David Hildenbrand <david@redhat.com>
---
arch/alpha/include/asm/io.h | 5 -----
arch/arm/include/asm/io.h | 5 -----
arch/hexagon/include/asm/io.h | 1 -
arch/ia64/include/asm/io.h | 1 -
arch/ia64/include/asm/uaccess.h | 18 ------------------
arch/m68k/include/asm/io_mm.h | 5 -----
arch/mips/include/asm/io.h | 5 -----
arch/parisc/include/asm/io.h | 5 -----
arch/powerpc/include/asm/io.h | 5 -----
arch/s390/include/asm/io.h | 5 -----
arch/sh/include/asm/io.h | 5 -----
arch/sparc/include/asm/io_64.h | 5 -----
include/asm-generic/io.h | 11 -----------
13 files changed, 76 deletions(-)
diff --git a/arch/alpha/include/asm/io.h b/arch/alpha/include/asm/io.h
index 1f6a909d1fa5..0fab5ac90775 100644
--- a/arch/alpha/include/asm/io.h
+++ b/arch/alpha/include/asm/io.h
@@ -602,11 +602,6 @@ extern void outsl (unsigned long port, const void *src, unsigned long count);
*/
#define xlate_dev_mem_ptr(p) __va(p)
-/*
- * Convert a virtual cached pointer to an uncached pointer
- */
-#define xlate_dev_kmem_ptr(p) p
-
#endif /* __KERNEL__ */
#endif /* __ALPHA_IO_H */
diff --git a/arch/arm/include/asm/io.h b/arch/arm/include/asm/io.h
index fc748122f1e0..f74944c6fe8d 100644
--- a/arch/arm/include/asm/io.h
+++ b/arch/arm/include/asm/io.h
@@ -430,11 +430,6 @@ extern void pci_iounmap(struct pci_dev *dev, void __iomem *addr);
*/
#define xlate_dev_mem_ptr(p) __va(p)
-/*
- * Convert a virtual cached pointer to an uncached pointer
- */
-#define xlate_dev_kmem_ptr(p) p
-
#include <asm-generic/io.h>
#ifdef CONFIG_MMU
diff --git a/arch/hexagon/include/asm/io.h b/arch/hexagon/include/asm/io.h
index bda2a9c2df78..c33241425a5c 100644
--- a/arch/hexagon/include/asm/io.h
+++ b/arch/hexagon/include/asm/io.h
@@ -64,7 +64,6 @@ static inline void *phys_to_virt(unsigned long address)
* convert a physical pointer to a virtual kernel pointer for
* /dev/mem access.
*/
-#define xlate_dev_kmem_ptr(p) __va(p)
#define xlate_dev_mem_ptr(p) __va(p)
/*
diff --git a/arch/ia64/include/asm/io.h b/arch/ia64/include/asm/io.h
index 3d666a11a2de..6d93b923b379 100644
--- a/arch/ia64/include/asm/io.h
+++ b/arch/ia64/include/asm/io.h
@@ -277,7 +277,6 @@ extern void memset_io(volatile void __iomem *s, int c, long n);
#define memcpy_fromio memcpy_fromio
#define memcpy_toio memcpy_toio
#define memset_io memset_io
-#define xlate_dev_kmem_ptr xlate_dev_kmem_ptr
#define xlate_dev_mem_ptr xlate_dev_mem_ptr
#include <asm-generic/io.h>
#undef PCI_IOBASE
diff --git a/arch/ia64/include/asm/uaccess.h b/arch/ia64/include/asm/uaccess.h
index 179243c3dfc7..e19d2dcc0ced 100644
--- a/arch/ia64/include/asm/uaccess.h
+++ b/arch/ia64/include/asm/uaccess.h
@@ -272,22 +272,4 @@ xlate_dev_mem_ptr(phys_addr_t p)
return ptr;
}
-/*
- * Convert a virtual cached kernel memory pointer to an uncached pointer
- */
-static __inline__ void *
-xlate_dev_kmem_ptr(void *p)
-{
- struct page *page;
- void *ptr;
-
- page = virt_to_page((unsigned long)p);
- if (PageUncached(page))
- ptr = (void *)__pa(p) + __IA64_UNCACHED_OFFSET;
- else
- ptr = p;
-
- return ptr;
-}
-
#endif /* _ASM_IA64_UACCESS_H */
diff --git a/arch/m68k/include/asm/io_mm.h b/arch/m68k/include/asm/io_mm.h
index 819f611dccf2..d41fa488453b 100644
--- a/arch/m68k/include/asm/io_mm.h
+++ b/arch/m68k/include/asm/io_mm.h
@@ -397,11 +397,6 @@ static inline void isa_delay(void)
*/
#define xlate_dev_mem_ptr(p) __va(p)
-/*
- * Convert a virtual cached pointer to an uncached pointer
- */
-#define xlate_dev_kmem_ptr(p) p
-
#define readb_relaxed(addr) readb(addr)
#define readw_relaxed(addr) readw(addr)
#define readl_relaxed(addr) readl(addr)
diff --git a/arch/mips/include/asm/io.h b/arch/mips/include/asm/io.h
index 78537aa23500..e6373e7ac892 100644
--- a/arch/mips/include/asm/io.h
+++ b/arch/mips/include/asm/io.h
@@ -552,11 +552,6 @@ extern void (*_dma_cache_inv)(unsigned long start, unsigned long size);
*/
#define xlate_dev_mem_ptr(p) __va(p)
-/*
- * Convert a virtual cached pointer to an uncached pointer
- */
-#define xlate_dev_kmem_ptr(p) p
-
void __ioread64_copy(void *to, const void __iomem *from, size_t count);
#endif /* _ASM_IO_H */
diff --git a/arch/parisc/include/asm/io.h b/arch/parisc/include/asm/io.h
index 8a11b8cf4719..0b5259102319 100644
--- a/arch/parisc/include/asm/io.h
+++ b/arch/parisc/include/asm/io.h
@@ -316,11 +316,6 @@ extern void iowrite64be(u64 val, void __iomem *addr);
*/
#define xlate_dev_mem_ptr(p) __va(p)
-/*
- * Convert a virtual cached pointer to an uncached pointer
- */
-#define xlate_dev_kmem_ptr(p) p
-
extern int devmem_is_allowed(unsigned long pfn);
#endif
diff --git a/arch/powerpc/include/asm/io.h b/arch/powerpc/include/asm/io.h
index 273edd208ec5..f130783c8301 100644
--- a/arch/powerpc/include/asm/io.h
+++ b/arch/powerpc/include/asm/io.h
@@ -662,11 +662,6 @@ static inline void name at \
*/
#define xlate_dev_mem_ptr(p) __va(p)
-/*
- * Convert a virtual cached pointer to an uncached pointer
- */
-#define xlate_dev_kmem_ptr(p) p
-
/*
* We don't do relaxed operations yet, at least not with this semantic
*/
diff --git a/arch/s390/include/asm/io.h b/arch/s390/include/asm/io.h
index 28664ee0abc1..e3882b012bfa 100644
--- a/arch/s390/include/asm/io.h
+++ b/arch/s390/include/asm/io.h
@@ -20,11 +20,6 @@ void *xlate_dev_mem_ptr(phys_addr_t phys);
#define unxlate_dev_mem_ptr unxlate_dev_mem_ptr
void unxlate_dev_mem_ptr(phys_addr_t phys, void *addr);
-/*
- * Convert a virtual cached pointer to an uncached pointer
- */
-#define xlate_dev_kmem_ptr(p) p
-
#define IO_SPACE_LIMIT 0
void __iomem *ioremap_prot(phys_addr_t addr, size_t size, unsigned long prot);
diff --git a/arch/sh/include/asm/io.h b/arch/sh/include/asm/io.h
index 6d5c6463bc07..cf9a3ec32406 100644
--- a/arch/sh/include/asm/io.h
+++ b/arch/sh/include/asm/io.h
@@ -283,11 +283,6 @@ static inline void __iomem *ioremap_prot(phys_addr_t offset, unsigned long size,
*/
#define xlate_dev_mem_ptr(p) __va(p)
-/*
- * Convert a virtual cached pointer to an uncached pointer
- */
-#define xlate_dev_kmem_ptr(p) p
-
#define ARCH_HAS_VALID_PHYS_ADDR_RANGE
int valid_phys_addr_range(phys_addr_t addr, size_t size);
int valid_mmap_phys_addr_range(unsigned long pfn, size_t size);
diff --git a/arch/sparc/include/asm/io_64.h b/arch/sparc/include/asm/io_64.h
index 9bb27e5c22f1..ff6fe387d78c 100644
--- a/arch/sparc/include/asm/io_64.h
+++ b/arch/sparc/include/asm/io_64.h
@@ -450,11 +450,6 @@ void sbus_set_sbus64(struct device *, int);
*/
#define xlate_dev_mem_ptr(p) __va(p)
-/*
- * Convert a virtual cached pointer to an uncached pointer
- */
-#define xlate_dev_kmem_ptr(p) p
-
#endif
#endif /* !(__SPARC64_IO_H) */
diff --git a/include/asm-generic/io.h b/include/asm-generic/io.h
index c6af40ce03be..33d4746b086f 100644
--- a/include/asm-generic/io.h
+++ b/include/asm-generic/io.h
@@ -1045,17 +1045,6 @@ static inline void pci_iounmap(struct pci_dev *dev, void __iomem *p)
#endif
#endif /* CONFIG_GENERIC_IOMAP */
-/*
- * Convert a virtual cached pointer to an uncached pointer
- */
-#ifndef xlate_dev_kmem_ptr
-#define xlate_dev_kmem_ptr xlate_dev_kmem_ptr
-static inline void *xlate_dev_kmem_ptr(void *addr)
-{
- return addr;
-}
-#endif
-
#ifndef xlate_dev_mem_ptr
#define xlate_dev_mem_ptr xlate_dev_mem_ptr
static inline void *xlate_dev_mem_ptr(phys_addr_t addr)
--
2.29.2
^ permalink raw reply related
* [PATCH v11 0/6] KASAN for powerpc64 radix
From: Daniel Axtens @ 2021-03-19 14:40 UTC (permalink / raw)
To: linux-kernel, linux-mm, linuxppc-dev, kasan-dev, christophe.leroy,
aneesh.kumar, bsingharora
Cc: Daniel Axtens
Building on the work of Christophe, Aneesh and Balbir, I've ported
KASAN to 64-bit Book3S kernels running on the Radix MMU.
v11 applies to next-20210317. I had hoped to have it apply to
powerpc/next but once again there are changes in the kasan core that
clash. Also, thanks to mpe for fixing a build break with KASAN off.
I'm not sure how best to progress this towards actually being merged
when it has impacts across subsystems. I'd appreciate any input. Maybe
the first four patches could go in via the kasan tree, that should
make things easier for powerpc in a future cycle?
v10 rebases on top of next-20210125, fixing things up to work on top
of the latest changes, and fixing some review comments from
Christophe. I have tested host and guest with 64k pages for this spin.
There is now only 1 failing KUnit test: kasan_global_oob - gcc puts
the ASAN init code in a section called '.init_array'. Powerpc64 module
loading code goes through and _renames_ any section beginning with
'.init' to begin with '_init' in order to avoid some complexities
around our 24-bit indirect jumps. This means it renames '.init_array'
to '_init_array', and the generic module loading code then fails to
recognise the section as a constructor and thus doesn't run it. This
hack dates back to 2003 and so I'm not going to try to unpick it in
this series. (I suspect this may have previously worked if the code
ended up in .ctors rather than .init_array but I don't keep my old
binaries around so I have no real way of checking.)
(The previously failing stack tests are now skipped due to more
accurate configuration settings.)
Details from v9: This is a significant reworking of the previous
versions. Instead of the previous approach which supported inline
instrumentation, this series provides only outline instrumentation.
To get around the problem of accessing the shadow region inside code we run
with translations off (in 'real mode'), we we restrict checking to when
translations are enabled. This is done via a new hook in the kasan core and
by excluding larger quantites of arch code from instrumentation. The upside
is that we no longer require that you be able to specify the amount of
physically contiguous memory on the system at compile time. Hopefully this
is a better trade-off. More details in patch 6.
kexec works. Both 64k and 4k pages work. Running as a KVM host works, but
nothing in arch/powerpc/kvm is instrumented. It's also potentially a bit
fragile - if any real mode code paths call out to instrumented code, things
will go boom.
Kind regards,
Daniel
Daniel Axtens (6):
kasan: allow an architecture to disable inline instrumentation
kasan: allow architectures to provide an outline readiness check
kasan: define and use MAX_PTRS_PER_* for early shadow tables
kasan: Document support on 32-bit powerpc
powerpc/mm/kasan: rename kasan_init_32.c to init_32.c
powerpc: Book3S 64-bit outline-only KASAN support
^ permalink raw reply
* [PATCH v11 1/6] kasan: allow an architecture to disable inline instrumentation
From: Daniel Axtens @ 2021-03-19 14:40 UTC (permalink / raw)
To: linux-kernel, linux-mm, linuxppc-dev, kasan-dev, christophe.leroy,
aneesh.kumar, bsingharora
Cc: Daniel Axtens
In-Reply-To: <20210319144058.772525-1-dja@axtens.net>
For annoying architectural reasons, it's very difficult to support inline
instrumentation on powerpc64.
Add a Kconfig flag to allow an arch to disable inline. (It's a bit
annoying to be 'backwards', but I'm not aware of any way to have
an arch force a symbol to be 'n', rather than 'y'.)
We also disable stack instrumentation in this case as it does things that
are functionally equivalent to inline instrumentation, namely adding
code that touches the shadow directly without going through a C helper.
Signed-off-by: Daniel Axtens <dja@axtens.net>
---
lib/Kconfig.kasan | 8 ++++++++
1 file changed, 8 insertions(+)
diff --git a/lib/Kconfig.kasan b/lib/Kconfig.kasan
index cffc2ebbf185..7e237dbb6df3 100644
--- a/lib/Kconfig.kasan
+++ b/lib/Kconfig.kasan
@@ -12,6 +12,9 @@ config HAVE_ARCH_KASAN_HW_TAGS
config HAVE_ARCH_KASAN_VMALLOC
bool
+config ARCH_DISABLE_KASAN_INLINE
+ def_bool n
+
config CC_HAS_KASAN_GENERIC
def_bool $(cc-option, -fsanitize=kernel-address)
@@ -130,6 +133,7 @@ config KASAN_OUTLINE
config KASAN_INLINE
bool "Inline instrumentation"
+ depends on !ARCH_DISABLE_KASAN_INLINE
help
Compiler directly inserts code checking shadow memory before
memory accesses. This is faster than outline (in some workloads
@@ -142,6 +146,7 @@ config KASAN_STACK
bool "Enable stack instrumentation (unsafe)" if CC_IS_CLANG && !COMPILE_TEST
depends on KASAN_GENERIC || KASAN_SW_TAGS
default y if CC_IS_GCC
+ depends on !ARCH_DISABLE_KASAN_INLINE
help
The LLVM stack address sanitizer has a know problem that
causes excessive stack usage in a lot of functions, see
@@ -154,6 +159,9 @@ config KASAN_STACK
but clang users can still enable it for builds without
CONFIG_COMPILE_TEST. On gcc it is assumed to always be safe
to use and enabled by default.
+ If the architecture disables inline instrumentation, this is
+ also disabled as it adds inline-style instrumentation that
+ is run unconditionally.
config KASAN_SW_TAGS_IDENTIFY
bool "Enable memory corruption identification"
--
2.27.0
^ permalink raw reply related
* [PATCH v11 2/6] kasan: allow architectures to provide an outline readiness check
From: Daniel Axtens @ 2021-03-19 14:40 UTC (permalink / raw)
To: linux-kernel, linux-mm, linuxppc-dev, kasan-dev, christophe.leroy,
aneesh.kumar, bsingharora
Cc: Aneesh Kumar K . V, Daniel Axtens
In-Reply-To: <20210319144058.772525-1-dja@axtens.net>
Allow architectures to define a kasan_arch_is_ready() hook that bails
out of any function that's about to touch the shadow unless the arch
says that it is ready for the memory to be accessed. This is fairly
uninvasive and should have a negligible performance penalty.
This will only work in outline mode, so an arch must specify
ARCH_DISABLE_KASAN_INLINE if it requires this.
Cc: Balbir Singh <bsingharora@gmail.com>
Cc: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Suggested-by: Christophe Leroy <christophe.leroy@csgroup.eu>
Signed-off-by: Daniel Axtens <dja@axtens.net>
--
I discuss the justfication for this later in the series. Also,
both previous RFCs for ppc64 - by 2 different people - have
needed this trick! See:
- https://lore.kernel.org/patchwork/patch/592820/ # ppc64 hash series
- https://patchwork.ozlabs.org/patch/795211/ # ppc radix series
---
include/linux/kasan.h | 4 ++++
mm/kasan/common.c | 4 ++++
mm/kasan/generic.c | 3 +++
mm/kasan/shadow.c | 4 ++++
4 files changed, 15 insertions(+)
diff --git a/include/linux/kasan.h b/include/linux/kasan.h
index 8b3b99d659b7..6bd8343f0033 100644
--- a/include/linux/kasan.h
+++ b/include/linux/kasan.h
@@ -23,6 +23,10 @@ struct kunit_kasan_expectation {
#endif
+#ifndef kasan_arch_is_ready
+static inline bool kasan_arch_is_ready(void) { return true; }
+#endif
+
#if defined(CONFIG_KASAN_GENERIC) || defined(CONFIG_KASAN_SW_TAGS)
#include <linux/pgtable.h>
diff --git a/mm/kasan/common.c b/mm/kasan/common.c
index 6bb87f2acd4e..f23a9e2dce9f 100644
--- a/mm/kasan/common.c
+++ b/mm/kasan/common.c
@@ -345,6 +345,10 @@ static inline bool ____kasan_slab_free(struct kmem_cache *cache, void *object,
if (unlikely(cache->flags & SLAB_TYPESAFE_BY_RCU))
return false;
+ /* We can't read the shadow byte if the arch isn't ready */
+ if (!kasan_arch_is_ready())
+ return false;
+
if (!kasan_byte_accessible(tagged_object)) {
kasan_report_invalid_free(tagged_object, ip);
return true;
diff --git a/mm/kasan/generic.c b/mm/kasan/generic.c
index 53cbf28859b5..c3f5ba7a294a 100644
--- a/mm/kasan/generic.c
+++ b/mm/kasan/generic.c
@@ -163,6 +163,9 @@ static __always_inline bool check_region_inline(unsigned long addr,
size_t size, bool write,
unsigned long ret_ip)
{
+ if (!kasan_arch_is_ready())
+ return true;
+
if (unlikely(size == 0))
return true;
diff --git a/mm/kasan/shadow.c b/mm/kasan/shadow.c
index 727ad4629173..1f650c521037 100644
--- a/mm/kasan/shadow.c
+++ b/mm/kasan/shadow.c
@@ -80,6 +80,10 @@ void kasan_poison(const void *addr, size_t size, u8 value, bool init)
*/
addr = kasan_reset_tag(addr);
+ /* Don't touch the shadow memory if arch isn't ready */
+ if (!kasan_arch_is_ready())
+ return;
+
/* Skip KFENCE memory if called explicitly outside of sl*b. */
if (is_kfence_address(addr))
return;
--
2.27.0
^ permalink raw reply related
* [PATCH v11 3/6] kasan: define and use MAX_PTRS_PER_* for early shadow tables
From: Daniel Axtens @ 2021-03-19 14:40 UTC (permalink / raw)
To: linux-kernel, linux-mm, linuxppc-dev, kasan-dev, christophe.leroy,
aneesh.kumar, bsingharora
Cc: Daniel Axtens
In-Reply-To: <20210319144058.772525-1-dja@axtens.net>
powerpc has a variable number of PTRS_PER_*, set at runtime based
on the MMU that the kernel is booted under.
This means the PTRS_PER_* are no longer constants, and therefore
breaks the build.
Define default MAX_PTRS_PER_*s in the same style as MAX_PTRS_PER_P4D.
As KASAN is the only user at the moment, just define them in the kasan
header, and have them default to PTRS_PER_* unless overridden in arch
code.
Suggested-by: Christophe Leroy <christophe.leroy@csgroup.eu>
Suggested-by: Balbir Singh <bsingharora@gmail.com>
Reviewed-by: Christophe Leroy <christophe.leroy@csgroup.eu>
Reviewed-by: Balbir Singh <bsingharora@gmail.com>
Signed-off-by: Daniel Axtens <dja@axtens.net>
---
include/linux/kasan.h | 18 +++++++++++++++---
mm/kasan/init.c | 6 +++---
2 files changed, 18 insertions(+), 6 deletions(-)
diff --git a/include/linux/kasan.h b/include/linux/kasan.h
index 6bd8343f0033..68cd6e55c872 100644
--- a/include/linux/kasan.h
+++ b/include/linux/kasan.h
@@ -44,10 +44,22 @@ static inline bool kasan_arch_is_ready(void) { return true; }
#define PTE_HWTABLE_PTRS 0
#endif
+#ifndef MAX_PTRS_PER_PTE
+#define MAX_PTRS_PER_PTE PTRS_PER_PTE
+#endif
+
+#ifndef MAX_PTRS_PER_PMD
+#define MAX_PTRS_PER_PMD PTRS_PER_PMD
+#endif
+
+#ifndef MAX_PTRS_PER_PUD
+#define MAX_PTRS_PER_PUD PTRS_PER_PUD
+#endif
+
extern unsigned char kasan_early_shadow_page[PAGE_SIZE];
-extern pte_t kasan_early_shadow_pte[PTRS_PER_PTE + PTE_HWTABLE_PTRS];
-extern pmd_t kasan_early_shadow_pmd[PTRS_PER_PMD];
-extern pud_t kasan_early_shadow_pud[PTRS_PER_PUD];
+extern pte_t kasan_early_shadow_pte[MAX_PTRS_PER_PTE + PTE_HWTABLE_PTRS];
+extern pmd_t kasan_early_shadow_pmd[MAX_PTRS_PER_PMD];
+extern pud_t kasan_early_shadow_pud[MAX_PTRS_PER_PUD];
extern p4d_t kasan_early_shadow_p4d[MAX_PTRS_PER_P4D];
int kasan_populate_early_shadow(const void *shadow_start,
diff --git a/mm/kasan/init.c b/mm/kasan/init.c
index c4605ac9837b..b4d822dff1fb 100644
--- a/mm/kasan/init.c
+++ b/mm/kasan/init.c
@@ -41,7 +41,7 @@ static inline bool kasan_p4d_table(pgd_t pgd)
}
#endif
#if CONFIG_PGTABLE_LEVELS > 3
-pud_t kasan_early_shadow_pud[PTRS_PER_PUD] __page_aligned_bss;
+pud_t kasan_early_shadow_pud[MAX_PTRS_PER_PUD] __page_aligned_bss;
static inline bool kasan_pud_table(p4d_t p4d)
{
return p4d_page(p4d) == virt_to_page(lm_alias(kasan_early_shadow_pud));
@@ -53,7 +53,7 @@ static inline bool kasan_pud_table(p4d_t p4d)
}
#endif
#if CONFIG_PGTABLE_LEVELS > 2
-pmd_t kasan_early_shadow_pmd[PTRS_PER_PMD] __page_aligned_bss;
+pmd_t kasan_early_shadow_pmd[MAX_PTRS_PER_PMD] __page_aligned_bss;
static inline bool kasan_pmd_table(pud_t pud)
{
return pud_page(pud) == virt_to_page(lm_alias(kasan_early_shadow_pmd));
@@ -64,7 +64,7 @@ static inline bool kasan_pmd_table(pud_t pud)
return false;
}
#endif
-pte_t kasan_early_shadow_pte[PTRS_PER_PTE + PTE_HWTABLE_PTRS]
+pte_t kasan_early_shadow_pte[MAX_PTRS_PER_PTE + PTE_HWTABLE_PTRS]
__page_aligned_bss;
static inline bool kasan_pte_table(pmd_t pmd)
--
2.27.0
^ permalink raw reply related
* [PATCH v11 4/6] kasan: Document support on 32-bit powerpc
From: Daniel Axtens @ 2021-03-19 14:40 UTC (permalink / raw)
To: linux-kernel, linux-mm, linuxppc-dev, kasan-dev, christophe.leroy,
aneesh.kumar, bsingharora
Cc: Daniel Axtens
In-Reply-To: <20210319144058.772525-1-dja@axtens.net>
KASAN is supported on 32-bit powerpc and the docs should reflect this.
Suggested-by: Christophe Leroy <christophe.leroy@csgroup.eu>
Reviewed-by: Christophe Leroy <christophe.leroy@csgroup.eu>
Signed-off-by: Daniel Axtens <dja@axtens.net>
---
Documentation/dev-tools/kasan.rst | 8 ++++++--
Documentation/powerpc/kasan.txt | 12 ++++++++++++
2 files changed, 18 insertions(+), 2 deletions(-)
create mode 100644 Documentation/powerpc/kasan.txt
diff --git a/Documentation/dev-tools/kasan.rst b/Documentation/dev-tools/kasan.rst
index a8c3e0cff88d..2cfd5d9068c0 100644
--- a/Documentation/dev-tools/kasan.rst
+++ b/Documentation/dev-tools/kasan.rst
@@ -36,7 +36,8 @@ Both software KASAN modes work with SLUB and SLAB memory allocators,
while the hardware tag-based KASAN currently only supports SLUB.
Currently, generic KASAN is supported for the x86_64, arm, arm64, xtensa, s390,
-and riscv architectures, and tag-based KASAN modes are supported only for arm64.
+and riscv architectures. It is also supported on 32-bit powerpc kernels.
+Tag-based KASAN modes are supported only for arm64.
Usage
-----
@@ -334,7 +335,10 @@ CONFIG_KASAN_VMALLOC
With ``CONFIG_KASAN_VMALLOC``, KASAN can cover vmalloc space at the
cost of greater memory usage. Currently, this is supported on x86,
-riscv, s390, and powerpc.
+riscv, s390, and 32-bit powerpc.
+
+It is optional, except on 32-bit powerpc kernels with module support,
+where it is required.
This works by hooking into vmalloc and vmap and dynamically
allocating real shadow memory to back the mappings.
diff --git a/Documentation/powerpc/kasan.txt b/Documentation/powerpc/kasan.txt
new file mode 100644
index 000000000000..26bb0e8bb18c
--- /dev/null
+++ b/Documentation/powerpc/kasan.txt
@@ -0,0 +1,12 @@
+KASAN is supported on powerpc on 32-bit only.
+
+32 bit support
+==============
+
+KASAN is supported on both hash and nohash MMUs on 32-bit.
+
+The shadow area sits at the top of the kernel virtual memory space above the
+fixmap area and occupies one eighth of the total kernel virtual memory space.
+
+Instrumentation of the vmalloc area is optional, unless built with modules,
+in which case it is required.
--
2.27.0
^ permalink raw reply related
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox