Skip to contentSkip to conversation

tip-queue 2016-02-16

PATCH New Borislav Petkov 2 participants 9 messages Feb 16, 2016 – Feb 17, 2016

Subsystems (3)
lore
Share & tools
lore

Borislav Petkov

author maintainer
#0 lore raw

cover tip-queue 2016-02-16

From: Borislav Petkov <[email protected]>

Hi,

just a bunch of small cleanups, the highlight being one less paravirt_enabled() user. :)

Please apply,
thanks.

Andrzej Hajda (1):
x86/microcode: Use kmemdup() rather than duplicating its
implementation

Boris Ostrovsky (1):
x86/microcode: Remove unnecessary paravirt_enabled check

Borislav Petkov (2):
ftrace: Kill ftrace_caller_end label
x86/msr: Document msr-index.h rule for addition

 arch/x86/include/asm/msr-index.h      |  7 ++++++-
 arch/x86/kernel/cpu/microcode/amd.c   |  4 +---
 arch/x86/kernel/cpu/microcode/core.c  |  2 +-
 arch/x86/kernel/cpu/microcode/intel.c |  4 +---
 arch/x86/kernel/ftrace.c              | 11 +++++------
 arch/x86/kernel/mcount_64.S           | 14 ++++++++------
 6 files changed, 22 insertions(+), 20 deletions(-)
Signature · 1 line
2.3.5
author maintainer
#1 lore raw

PATCH 1/4 x86/microcode: Remove unnecessary paravirt_enabled check

From: Boris Ostrovsky <[email protected]>

Commit

a18a0f6850d4 ("x86, microcode: Don't initialize microcode code on paravirt")

added a paravirt test in microcode_init(), primarily to avoid making mc_bp_resume()->load_ucode_ap()->check_loader_disabled_ap() calls because on 32-bit kernels this callchain ends up using __pa_nodebug() macro which is invalid for Xen PV guests.

A subsequent commit

fbae4ba8c4a3 ("x86, microcode: Reload microcode on resume")

eliminated this callchain thus making a18a0f6850d4 unnecessary.

---
 arch/x86/kernel/cpu/microcode/core.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
arch/x86/kernel/cpu/microcode/core.c +1 -1
index faec7120c508..86ec92a0bbbe 100644
@@ -630,7 +630,7 @@ int __init microcode_init(void)
 	struct cpuinfo_x86 *c = &boot_cpu_data;
 	int error;
 
-	if (paravirt_enabled() || dis_ucode_ldr)
+	if (dis_ucode_ldr)
 		return -EINVAL;
 
 	if (c->x86_vendor == X86_VENDOR_INTEL)
Signature · 1 line
2.3.5
#2 lore raw

[tip:x86/microcode] x86/microcode: Remove unnecessary paravirt_enabled check

x86/microcode: Remove unnecessary paravirt_enabled check
Commit-ID:  84aba677f009e20185aea322563389ad56e0ef7e
Gitweb:     http://git.kernel.org/tip/84aba677f009e20185aea322563389ad56e0ef7e
Author:     Boris Ostrovsky <[email protected]>
AuthorDate: Tue, 16 Feb 2016 09:43:19 +0100
Committer:  Ingo Molnar <[email protected]>
CommitDate: Wed, 17 Feb 2016 08:46:07 +0100

x86/microcode: Remove unnecessary paravirt_enabled check

Commit:

a18a0f6850d4 ("x86, microcode: Don't initialize microcode code on paravirt")

added a paravirt test in microcode_init(), primarily to avoid making mc_bp_resume()->load_ucode_ap()->check_loader_disabled_ap() calls because on 32-bit kernels this callchain ends up using __pa_nodebug() macro which is invalid for Xen PV guests.

A subsequent commit:

fbae4ba8c4a3 ("x86, microcode: Reload microcode on resume")

eliminated this callchain thus making a18a0f6850d4 unnecessary.

---
 arch/x86/kernel/cpu/microcode/core.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
arch/x86/kernel/cpu/microcode/core.c +1 -1
index cea8552..ac360bf 100644
@@ -623,7 +623,7 @@ int __init microcode_init(void)
 	struct cpuinfo_x86 *c = &boot_cpu_data;
 	int error;
 
-	if (paravirt_enabled() || dis_ucode_ldr)
+	if (dis_ucode_ldr)
 		return -EINVAL;
 
 	if (c->x86_vendor == X86_VENDOR_INTEL)
author maintainer
#3 lore raw

PATCH 2/4 x86/microcode: Use kmemdup() rather than duplicating its implementation

In reply to Borislav Petkov

From: Andrzej Hajda <[email protected]>

The patch was generated using fixed coccinelle semantic patch scripts/coccinelle/api/memdup.cocci.

Cc Bartlomiej Zolnierkiewicz <[email protected]>
Cc H. Peter Anvin <[email protected]>
Cc Linus Torvalds <[email protected]>
Cc Marek Szyprowski <[email protected]>
Cc Peter Zijlstra <[email protected]>
Cc Thomas Gleixner <[email protected]>
---
 arch/x86/kernel/cpu/microcode/amd.c   | 4 +---
 arch/x86/kernel/cpu/microcode/intel.c | 4 +---
 2 files changed, 2 insertions(+), 6 deletions(-)
arch/x86/kernel/cpu/microcode/amd.c +1 -3
index 2233f8a76615..34b391bb71c5 100644
@@ -793,15 +793,13 @@ static int verify_and_add_patch(u8 family, u8 *fw, unsigned int leftover)
 		return -EINVAL;
 	}
 
-	patch->data = kzalloc(patch_size, GFP_KERNEL);
+	patch->data = kmemdup(fw + SECTION_HDR_SIZE, patch_size, GFP_KERNEL);
 	if (!patch->data) {
 		pr_err("Patch data allocation failure.\n");
 		kfree(patch);
 		return -EINVAL;
 	}
 
-	/* All looks ok, copy patch... */
-	memcpy(patch->data, fw + SECTION_HDR_SIZE, patch_size);
 	INIT_LIST_HEAD(&patch->plist);
 	patch->patch_id  = mc_hdr->patch_id;
 	patch->equiv_cpu = proc_id;
arch/x86/kernel/cpu/microcode/intel.c +1 -3
index ee81c544ee0d..b7efe1f02f01 100644
@@ -207,13 +207,11 @@ save_microcode(struct mc_saved_data *mc_saved_data,
 		mc_hdr = &mc->hdr;
 		size   = get_totalsize(mc_hdr);
 
-		saved_ptr[i] = kmalloc(size, GFP_KERNEL);
+		saved_ptr[i] = kmemdup(mc, size, GFP_KERNEL);
 		if (!saved_ptr[i]) {
 			ret = -ENOMEM;
 			goto err;
 		}
-
-		memcpy(saved_ptr[i], mc, size);
 	}
 
 	/*
Signature · 1 line
2.3.5
#4 lore raw

[tip:x86/microcode] x86/microcode: Use kmemdup() rather than duplicating its implementation

x86/microcode: Use kmemdup() rather than duplicating its implementation
Commit-ID:  9cc6f743c7724eb9abaf27904194c169db85dd31
Gitweb:     http://git.kernel.org/tip/9cc6f743c7724eb9abaf27904194c169db85dd31
Author:     Andrzej Hajda <[email protected]>
AuthorDate: Tue, 16 Feb 2016 09:43:20 +0100
Committer:  Ingo Molnar <[email protected]>
CommitDate: Wed, 17 Feb 2016 08:46:08 +0100

x86/microcode: Use kmemdup() rather than duplicating its implementation

The patch was generated using fixed coccinelle semantic patch scripts/coccinelle/api/memdup.cocci.

---
 arch/x86/kernel/cpu/microcode/amd.c   | 4 +---
 arch/x86/kernel/cpu/microcode/intel.c | 4 +---
 2 files changed, 2 insertions(+), 6 deletions(-)
arch/x86/kernel/cpu/microcode/amd.c +1 -3
index f66cbfe..e397fc1 100644
@@ -788,15 +788,13 @@ static int verify_and_add_patch(u8 family, u8 *fw, unsigned int leftover)
 		return -EINVAL;
 	}
 
-	patch->data = kzalloc(patch_size, GFP_KERNEL);
+	patch->data = kmemdup(fw + SECTION_HDR_SIZE, patch_size, GFP_KERNEL);
 	if (!patch->data) {
 		pr_err("Patch data allocation failure.\n");
 		kfree(patch);
 		return -EINVAL;
 	}
 
-	/* All looks ok, copy patch... */
-	memcpy(patch->data, fw + SECTION_HDR_SIZE, patch_size);
 	INIT_LIST_HEAD(&patch->plist);
 	patch->patch_id  = mc_hdr->patch_id;
 	patch->equiv_cpu = proc_id;
arch/x86/kernel/cpu/microcode/intel.c +1 -3
index cb397947..cbb3cf0 100644
@@ -210,13 +210,11 @@ save_microcode(struct mc_saved_data *mcs,
 		mc_hdr = &mc->hdr;
 		size   = get_totalsize(mc_hdr);
 
-		saved_ptr[i] = kmalloc(size, GFP_KERNEL);
+		saved_ptr[i] = kmemdup(mc, size, GFP_KERNEL);
 		if (!saved_ptr[i]) {
 			ret = -ENOMEM;
 			goto err;
 		}
-
-		memcpy(saved_ptr[i], mc, size);
 	}
 
 	/*
author maintainer
#5 lore raw

PATCH 3/4 ftrace: Kill ftrace_caller_end label

In reply to Borislav Petkov

From: Borislav Petkov <[email protected]>

One of ftrace_caller_end and ftrace_return is redundant so unify them. Rename ftrace_return to ftrace_epilogue to mean that everything after that label represents, like an afterword, work which happens *after* the ftrace call, e.g., the function graph tracer for one.

Steve wants this to rather mean "[a]n event which reflects meaningfully on a recently ended conflict or struggle." I can imagine that ftrace can be a struggle sometimes.

Anyway, beef up the comment about the code contents and layout before ftrace_epilogue label.

---
 arch/x86/kernel/ftrace.c    | 11 +++++------
 arch/x86/kernel/mcount_64.S | 14 ++++++++------
 2 files changed, 13 insertions(+), 12 deletions(-)
arch/x86/kernel/ftrace.c +5 -6
index 29408d6d6626..04f9641e0cb6 100644
@@ -697,9 +697,8 @@ static inline void tramp_free(void *tramp) { }
 #endif
 
 /* Defined as markers to the end of the ftrace default trampolines */
-extern void ftrace_caller_end(void);
 extern void ftrace_regs_caller_end(void);
-extern void ftrace_return(void);
+extern void ftrace_epilogue(void);
 extern void ftrace_caller_op_ptr(void);
 extern void ftrace_regs_caller_op_ptr(void);
 
@@ -746,7 +745,7 @@ create_trampoline(struct ftrace_ops *ops, unsigned int *tramp_size)
 		op_offset = (unsigned long)ftrace_regs_caller_op_ptr;
 	} else {
 		start_offset = (unsigned long)ftrace_caller;
-		end_offset = (unsigned long)ftrace_caller_end;
+		end_offset = (unsigned long)ftrace_epilogue;
 		op_offset = (unsigned long)ftrace_caller_op_ptr;
 	}
 
@@ -754,7 +753,7 @@ create_trampoline(struct ftrace_ops *ops, unsigned int *tramp_size)
 
 	/*
 	 * Allocate enough size to store the ftrace_caller code,
-	 * the jmp to ftrace_return, as well as the address of
+	 * the jmp to ftrace_epilogue, as well as the address of
 	 * the ftrace_ops this trampoline is used for.
 	 */
 	trampoline = alloc_tramp(size + MCOUNT_INSN_SIZE + sizeof(void *));
@@ -772,8 +771,8 @@ create_trampoline(struct ftrace_ops *ops, unsigned int *tramp_size)
 
 	ip = (unsigned long)trampoline + size;
 
-	/* The trampoline ends with a jmp to ftrace_return */
-	jmp = ftrace_jmp_replace(ip, (unsigned long)ftrace_return);
+	/* The trampoline ends with a jmp to ftrace_epilogue */
+	jmp = ftrace_jmp_replace(ip, (unsigned long)ftrace_epilogue);
 	memcpy(trampoline + size, jmp, MCOUNT_INSN_SIZE);
 
 	/*
arch/x86/kernel/mcount_64.S +8 -6
index 87e1762e2bca..ed48a9f465f8 100644
@@ -168,12 +168,14 @@ GLOBAL(ftrace_call)
 	restore_mcount_regs
 
 	/*
-	 * The copied trampoline must call ftrace_return as it
+	 * The copied trampoline must call ftrace_epilogue as it
 	 * still may need to call the function graph tracer.
+	 *
+	 * The code up to this label is copied into trampolines so
+	 * think twice before adding any new code or changing the
+	 * layout here.
 	 */
-GLOBAL(ftrace_caller_end)
-
-GLOBAL(ftrace_return)
+GLOBAL(ftrace_epilogue)
 
 #ifdef CONFIG_FUNCTION_GRAPH_TRACER
 GLOBAL(ftrace_graph_call)
@@ -244,14 +246,14 @@ GLOBAL(ftrace_regs_call)
 	popfq
 
 	/*
-	 * As this jmp to ftrace_return can be a short jump
+	 * As this jmp to ftrace_epilogue can be a short jump
 	 * it must not be copied into the trampoline.
 	 * The trampoline will add the code to jump
 	 * to the return.
 	 */
 GLOBAL(ftrace_regs_caller_end)
 
-	jmp ftrace_return
+	jmp ftrace_epilogue
 
 END(ftrace_regs_caller)
 
Signature · 1 line
2.3.5
#6 lore raw

[tip:x86/asm] x86/ftrace, x86/asm: Kill ftrace_caller_end label

x86/ftrace, x86/asm: Kill ftrace_caller_end label
Commit-ID:  f1b92bb6b5a4e17b508f128b084fa00e0eda590c
Gitweb:     http://git.kernel.org/tip/f1b92bb6b5a4e17b508f128b084fa00e0eda590c
Author:     Borislav Petkov <[email protected]>
AuthorDate: Tue, 16 Feb 2016 09:43:21 +0100
Committer:  Ingo Molnar <[email protected]>
CommitDate: Wed, 17 Feb 2016 08:47:22 +0100

x86/ftrace, x86/asm: Kill ftrace_caller_end label

One of ftrace_caller_end and ftrace_return is redundant so unify them. Rename ftrace_return to ftrace_epilogue to mean that everything after that label represents, like an afterword, work which happens *after* the ftrace call, e.g., the function graph tracer for one.

Steve wants this to rather mean "[a]n event which reflects meaningfully on a recently ended conflict or struggle." I can imagine that ftrace can be a struggle sometimes.

Anyway, beef up the comment about the code contents and layout before ftrace_epilogue label.

Cc Andy Lutomirski <[email protected]>
Cc Borislav Petkov <[email protected]>
Cc Brian Gerst <[email protected]>
Cc Denys Vlasenko <[email protected]>
Cc H. Peter Anvin <[email protected]>
Cc Linus Torvalds <[email protected]>
Cc Peter Zijlstra <[email protected]>
Cc Thomas Gleixner <[email protected]>
---
 arch/x86/kernel/ftrace.c    | 11 +++++------
 arch/x86/kernel/mcount_64.S | 14 ++++++++------
 2 files changed, 13 insertions(+), 12 deletions(-)
arch/x86/kernel/ftrace.c +5 -6
index 29408d6..04f9641e 100644
@@ -697,9 +697,8 @@ static inline void tramp_free(void *tramp) { }
 #endif
 
 /* Defined as markers to the end of the ftrace default trampolines */
-extern void ftrace_caller_end(void);
 extern void ftrace_regs_caller_end(void);
-extern void ftrace_return(void);
+extern void ftrace_epilogue(void);
 extern void ftrace_caller_op_ptr(void);
 extern void ftrace_regs_caller_op_ptr(void);
 
@@ -746,7 +745,7 @@ create_trampoline(struct ftrace_ops *ops, unsigned int *tramp_size)
 		op_offset = (unsigned long)ftrace_regs_caller_op_ptr;
 	} else {
 		start_offset = (unsigned long)ftrace_caller;
-		end_offset = (unsigned long)ftrace_caller_end;
+		end_offset = (unsigned long)ftrace_epilogue;
 		op_offset = (unsigned long)ftrace_caller_op_ptr;
 	}
 
@@ -754,7 +753,7 @@ create_trampoline(struct ftrace_ops *ops, unsigned int *tramp_size)
 
 	/*
 	 * Allocate enough size to store the ftrace_caller code,
-	 * the jmp to ftrace_return, as well as the address of
+	 * the jmp to ftrace_epilogue, as well as the address of
 	 * the ftrace_ops this trampoline is used for.
 	 */
 	trampoline = alloc_tramp(size + MCOUNT_INSN_SIZE + sizeof(void *));
@@ -772,8 +771,8 @@ create_trampoline(struct ftrace_ops *ops, unsigned int *tramp_size)
 
 	ip = (unsigned long)trampoline + size;
 
-	/* The trampoline ends with a jmp to ftrace_return */
-	jmp = ftrace_jmp_replace(ip, (unsigned long)ftrace_return);
+	/* The trampoline ends with a jmp to ftrace_epilogue */
+	jmp = ftrace_jmp_replace(ip, (unsigned long)ftrace_epilogue);
 	memcpy(trampoline + size, jmp, MCOUNT_INSN_SIZE);
 
 	/*
arch/x86/kernel/mcount_64.S +8 -6
index 87e1762..ed48a9f 100644
@@ -168,12 +168,14 @@ GLOBAL(ftrace_call)
 	restore_mcount_regs
 
 	/*
-	 * The copied trampoline must call ftrace_return as it
+	 * The copied trampoline must call ftrace_epilogue as it
 	 * still may need to call the function graph tracer.
+	 *
+	 * The code up to this label is copied into trampolines so
+	 * think twice before adding any new code or changing the
+	 * layout here.
 	 */
-GLOBAL(ftrace_caller_end)
-
-GLOBAL(ftrace_return)
+GLOBAL(ftrace_epilogue)
 
 #ifdef CONFIG_FUNCTION_GRAPH_TRACER
 GLOBAL(ftrace_graph_call)
@@ -244,14 +246,14 @@ GLOBAL(ftrace_regs_call)
 	popfq
 
 	/*
-	 * As this jmp to ftrace_return can be a short jump
+	 * As this jmp to ftrace_epilogue can be a short jump
 	 * it must not be copied into the trampoline.
 	 * The trampoline will add the code to jump
 	 * to the return.
 	 */
 GLOBAL(ftrace_regs_caller_end)
 
-	jmp ftrace_return
+	jmp ftrace_epilogue
 
 END(ftrace_regs_caller)
 
author maintainer
#7 lore raw

PATCH 4/4 x86/msr: Document msr-index.h rule for addition

In reply to Borislav Petkov

From: Borislav Petkov <[email protected]>

In order to keep this file's size sensible and not cause too much unnecessary churn, make the rule explicit - similar to pci_ids.h - that only MSRs which are used in multiple compilation units, should get added to it.

---
 arch/x86/include/asm/msr-index.h | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)
arch/x86/include/asm/msr-index.h +6 -1
index b05402ef3b84..984ab75bf621 100644
@@ -1,7 +1,12 @@
 #ifndef _ASM_X86_MSR_INDEX_H
 #define _ASM_X86_MSR_INDEX_H
 
-/* CPU model specific register (MSR) numbers */
+/*
+ * CPU model specific register (MSR) numbers.
+ *
+ * Do not add new entries to this file unless the definitions are shared
+ * between multiple compilation units.
+ */
 
 /* x86-64 specific MSRs */
 #define MSR_EFER		0xc0000080 /* extended feature register */
Signature · 1 line
2.3.5
#8 lore raw

[tip:x86/asm] x86/msr: Document msr-index.h rule for addition

x86/msr: Document msr-index.h rule for addition
Commit-ID:  053080a9d1c8cf1950115ad92ce94242ebc5f25c
Gitweb:     http://git.kernel.org/tip/053080a9d1c8cf1950115ad92ce94242ebc5f25c
Author:     Borislav Petkov <[email protected]>
AuthorDate: Tue, 16 Feb 2016 09:43:22 +0100
Committer:  Ingo Molnar <[email protected]>
CommitDate: Wed, 17 Feb 2016 08:47:55 +0100

x86/msr: Document msr-index.h rule for addition

In order to keep this file's size sensible and not cause too much unnecessary churn, make the rule explicit - similar to pci_ids.h - that only MSRs which are used in multiple compilation units, should get added to it.

---
 arch/x86/include/asm/msr-index.h | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)
arch/x86/include/asm/msr-index.h +6 -1
index b05402e..984ab75 100644
@@ -1,7 +1,12 @@
 #ifndef _ASM_X86_MSR_INDEX_H
 #define _ASM_X86_MSR_INDEX_H
 
-/* CPU model specific register (MSR) numbers */
+/*
+ * CPU model specific register (MSR) numbers.
+ *
+ * Do not add new entries to this file unless the definitions are shared
+ * between multiple compilation units.
+ */
 
 /* x86-64 specific MSRs */
 #define MSR_EFER		0xc0000080 /* extended feature register */