Skip to contentSkip to conversation

add support for never printing hashed addresses

PATCH Superseded Timur Tabi 5 participants 15 messages Feb 10, 2021 – Feb 12, 2021

Subsystems (6)

v3 of 4 Superseded Latest is v4 →

lore
Share & tools
lore

Timur Tabi

author
#0 lore raw

v3 cover add support for never printing hashed addresses

[The list of email addresses on CC: is getting quite lengthy, so I hope I've included everyone.]

Although hashing addresses printed via printk does make the kernel more secure, it interferes with debugging, especially with some functions like print_hex_dump() which always uses hashed addresses.

To avoid having to choose between %p and %px, it's easier to add a kernel command line that treats all %p as %px. This encourages developers to use %p more without making debugging more difficult.

Patches #1 and #2 upgrade the kselftest framework so that it can report on tests that were skipped outright. This is needed for the test_printf module which will now skip %p hashing tests if hashing is disabled.

Patch #2 upgrades the printf library to check the command line. It also updates test_printf().

Full series:

Timur Tabi (3):
[v3] lib: use KSTM_MODULE_GLOBALS macro in kselftest drivers [v3] kselftest: add support for skipped tests
[v3] lib/vsprintf: debug_never_hash_pointers prints all addresses as
unhashed

 .../admin-guide/kernel-parameters.txt         | 15 ++++++++
 lib/test_bitmap.c                             |  3 +-
 lib/test_printf.c                             | 12 +++++-
 lib/vsprintf.c                                | 38 ++++++++++++++++++-
 tools/testing/selftests/kselftest_module.h    | 18 ++++++---
 5 files changed, 74 insertions(+), 12 deletions(-)
Signature · 1 line
2.25.1
author
#1 lore raw

PATCH v3 1/3 lib: use KSTM_MODULE_GLOBALS macro in kselftest drivers

Instead of defining the total/failed test counters manually, test drivers that are clients of kselftest should use the macro created for this purpose.

---
 lib/test_bitmap.c | 3 +--
 lib/test_printf.c | 4 ++--
 2 files changed, 3 insertions(+), 4 deletions(-)
lib/test_bitmap.c +1 -2
index 4425a1dd4ef1..0ea0e8258f14 100644
@@ -16,8 +16,7 @@
 
 #include "../tools/testing/selftests/kselftest_module.h"
 
-static unsigned total_tests __initdata;
-static unsigned failed_tests __initdata;
+KSTM_MODULE_GLOBALS();
 
 static char pbl_buffer[PAGE_SIZE] __initdata;
 
lib/test_printf.c +2 -2
index 7ac87f18a10f..ad2bcfa8caa1 100644
@@ -30,8 +30,8 @@
 #define PAD_SIZE 16
 #define FILL_CHAR '$'
 
-static unsigned total_tests __initdata;
-static unsigned failed_tests __initdata;
+KSTM_MODULE_GLOBALS();
+
 static char *test_buffer __initdata;
 static char *alloced_buffer __initdata;
 
Signature · 1 line
2.25.1
author
#2 lore raw

PATCH v3 2/3 kselftest: add support for skipped tests

In reply to Timur Tabi

Update the kselftest framework to allow client drivers to specify that some tests were skipped.

---
 tools/testing/selftests/kselftest_module.h | 18 ++++++++++++------
 1 file changed, 12 insertions(+), 6 deletions(-)
tools/testing/selftests/kselftest_module.h +12 -6
index e8eafaf0941a..e2ea41de3f35 100644
@@ -11,7 +11,8 @@
 
 #define KSTM_MODULE_GLOBALS()			\
 static unsigned int total_tests __initdata;	\
-static unsigned int failed_tests __initdata
+static unsigned int failed_tests __initdata;	\
+static unsigned int skipped_tests __initdata
 
 #define KSTM_CHECK_ZERO(x) do {						\
 	total_tests++;							\
@@ -21,11 +22,16 @@ static unsigned int failed_tests __initdata
 	}								\
 } while (0)
 
-static inline int kstm_report(unsigned int total_tests, unsigned int failed_tests)
+static inline int kstm_report(unsigned int total_tests, unsigned int failed_tests,
+			      unsigned int skipped_tests)
 {
-	if (failed_tests == 0)
-		pr_info("all %u tests passed\n", total_tests);
-	else
+	if (failed_tests == 0) {
+		if (skipped_tests) {
+			pr_info("skipped %u tests\n", skipped_tests);
+			pr_info("remaining %u tests passed\n", total_tests);
+		} else
+			pr_info("all %u tests passed\n", total_tests);
+	} else
 		pr_warn("failed %u out of %u tests\n", failed_tests, total_tests);
 
 	return failed_tests ? -EINVAL : 0;
@@ -36,7 +42,7 @@ static int __init __module##_init(void)			\
 {							\
 	pr_info("loaded.\n");				\
 	selftest();					\
-	return kstm_report(total_tests, failed_tests);	\
+	return kstm_report(total_tests, failed_tests, skipped_tests);	\
 }							\
 static void __exit __module##_exit(void)		\
 {							\
Signature · 1 line
2.25.1
author
#4 lore raw

PATCH v3 3/3 lib/vsprintf: debug_never_hash_pointers prints all addresses as unhashed

In reply to Timur Tabi

If the debug_never_hash_pointers command line parameter is set, then printk("%p") will print pointers as unhashed, which is useful for debugging purposes. This also applies to any function that uses vsprintf, such as print_hex_dump() and seq_buf_printf().

A large warning message is displayed if this option is enabled. Unhashed pointers expose kernel addresses, which can be a security risk.

Also update test_printf to skip the hashed pointer tests if the command-line option is set.

---
 .../admin-guide/kernel-parameters.txt         | 15 ++++++++
 lib/test_printf.c                             |  8 ++++
 lib/vsprintf.c                                | 38 ++++++++++++++++++-
 3 files changed, 59 insertions(+), 2 deletions(-)
Documentation/admin-guide/kernel-parameters.txt +15 -0
index a10b545c2070..2a97e787f49c 100644
@@ -810,6 +810,21 @@
 			1 will print _a lot_ more information - normally
 			only useful to kernel developers.
 
+	debug_never_hash_pointers
+			Force pointers printed to the console or buffers to be
+			unhashed.  By default, when a pointer is printed via %p
+			format string, that pointer is "hashed", i.e. obscured
+			by hashing the pointer value.  This is a security feature
+			that hides actual kernel addresses from unprivileged
+			users, but it also makes debugging the kernel more
+			difficult since unequal pointers can no longer be
+			compared.  However, if this command-line option is
+			specified, then all normal pointers will have their true
+			value printed.  Pointers printed via %pK may still be
+			hashed.  This option should only be specified when
+			debugging the kernel.  Please do not use on production
+			kernels.
+
 	debug_objects	[KNL] Enable object debugging
 
 	no_debug_objects
lib/test_printf.c +8 -0
index ad2bcfa8caa1..b0b62d76e598 100644
@@ -35,6 +35,8 @@ KSTM_MODULE_GLOBALS();
 static char *test_buffer __initdata;
 static char *alloced_buffer __initdata;
 
+extern bool debug_never_hash_pointers;
+
 static int __printf(4, 0) __init
 do_test(int bufsize, const char *expect, int elen,
 	const char *fmt, va_list ap)
@@ -301,6 +303,12 @@ plain(void)
 {
 	int err;
 
+	if (debug_never_hash_pointers) {
+		pr_warn("skipping plain 'p' tests");
+		skipped_tests += 2;
+		return;
+	}
+
 	err = plain_hash();
 	if (err) {
 		pr_warn("plain 'p' does not appear to be hashed\n");
lib/vsprintf.c +36 -2
index 3b53c73580c5..b4e07ecb1cb2 100644
@@ -2090,6 +2090,34 @@ char *fwnode_string(char *buf, char *end, struct fwnode_handle *fwnode,
 	return widen_string(buf, buf - buf_start, end, spec);
 }
 
+/* Disable pointer hashing if requested */
+bool debug_never_hash_pointers __ro_after_init;
+EXPORT_SYMBOL_GPL(debug_never_hash_pointers);
+
+static int __init debug_never_hash_pointers_enable(char *str)
+{
+	debug_never_hash_pointers = true;
+
+	pr_warn("**********************************************************\n");
+	pr_warn("**   NOTICE NOTICE NOTICE NOTICE NOTICE NOTICE NOTICE   **\n");
+	pr_warn("**                                                      **\n");
+	pr_warn("** All pointers that are printed to the console will    **\n");
+	pr_warn("** be printed as unhashed.                              **\n");
+	pr_warn("**                                                      **\n");
+	pr_warn("** Kernel memory addresses are exposed, which may       **\n");
+	pr_warn("** reduce the security of your system.                  **\n");
+	pr_warn("**                                                      **\n");
+	pr_warn("** If you see this message and you are not debugging    **\n");
+	pr_warn("** the kernel, report this immediately to your system   **\n");
+	pr_warn("** administrator!                                       **\n");
+	pr_warn("**                                                      **\n");
+	pr_warn("**   NOTICE NOTICE NOTICE NOTICE NOTICE NOTICE NOTICE   **\n");
+	pr_warn("**********************************************************\n");
+
+	return 0;
+}
+early_param("debug_never_hash_pointers", debug_never_hash_pointers_enable);
+
 /*
  * Show a '%p' thing.  A kernel extension is that the '%p' is followed
  * by an extra set of alphanumeric characters that are extended format
@@ -2297,8 +2325,14 @@ char *pointer(const char *fmt, char *buf, char *end, void *ptr,
 		}
 	}
 
-	/* default is to _not_ leak addresses, hash before printing */
-	return ptr_to_id(buf, end, ptr, spec);
+	/*
+	 * default is to _not_ leak addresses, so hash before printing,
+	 * unless debug_never_hash_pointers is specified on the command line.
+	 */
+	if (unlikely(debug_never_hash_pointers))
+		return pointer_string(buf, end, ptr, spec);
+	else
+		return ptr_to_id(buf, end, ptr, spec);
 }
 
 /*
Signature · 1 line
2.25.1
#5 lore raw

Re: [PATCH 3/3] [v3] lib/vsprintf: debug_never_hash_pointers prints all addresses as unhashed

Hi!

Quoted 1 line

If the debug_never_hash_pointers command line parameter is set, then

Can we make this something shorter? Clearly you don't want people placing this in their grub config, so they'll be most likely typing this a lot...

debug_pointers or debug_ptrs would be better.

Thanks,
Pavel

Signature · 1 line
http://www.livejournal.com/~pavelmachek
author
#6 lore raw

Re: [PATCH 3/3] [v3] lib/vsprintf: debug_never_hash_pointers prints all addresses as unhashed

Quoted 5 lines · Pavel Machek

Can we make this something shorter? Clearly you don't want people placing this in their grub config, so they'll be most likely typing this a lot...

debug_pointers or debug_ptrs would be better.

dbg_unhash_ptrs? "debug_ptrs" is too vague IMHO, and I want to keep the word "hash" somewhere there to indicate exactly what's happening.

#7 lore raw

Re: [PATCH 3/3] [v3] lib/vsprintf: debug_never_hash_pointers prints all addresses as unhashed

Quoted 11 lines · Timur Tabi
Quoted 5 lines · Pavel Machek

Can we make this something shorter? Clearly you don't want people placing this in their grub config, so they'll be most likely typing this a lot...

debug_pointers or debug_ptrs would be better.

dbg_unhash_ptrs? "debug_ptrs" is too vague IMHO, and I want to keep the word "hash" somewhere there to indicate exactly what's happening.

no_hash_pointers ?

maintainer
#8 lore raw

Re: [PATCH 3/3] [v3] lib/vsprintf: debug_never_hash_pointers prints all addresses as unhashed

Quoted 8 lines · Matthew Wilcox
Quoted 7 lines · Timur Tabi
Quoted 4 lines · Pavel Machek

Can we make this something shorter? Clearly you don't want people placing this in their grub config, so they'll be most likely typing this a lot...

Quoted 6 lines
Quoted 4 lines
Quoted 1 line

debug_pointers or debug_ptrs would be better.

dbg_unhash_ptrs? "debug_ptrs" is too vague IMHO, and I want to keep the word "hash" somewhere there to indicate exactly what's happening.

no_hash_pointers ?

I am fine with this.

I am still a bit scared of a bikeshedng. But AFAIK, Mathew was most active on proposing clear names. So, when he is fine with this...

Anyway, we should use the same name also for the variable.

Best Regards,
Petr

author
#9 lore raw

Re: [PATCH 3/3] [v3] lib/vsprintf: debug_never_hash_pointers prints all addresses as unhashed

Quoted 7 lines · Petr Mladek
Quoted 1 line

no_hash_pointers ?

I am fine with this.

I am still a bit scared of a bikeshedng. But AFAIK, Mathew was most active on proposing clear names. So, when he is fine with this...

Anyway, we should use the same name also for the variable.

Ok, unless there are any objections, I will change the parameter and variable to "no_hash_pointers" in v4, which I will send out later today.

I hope this patch set gets accepted for 5.12.

maintainer
#10 lore raw

Re: [PATCH 3/3] [v3] lib/vsprintf: debug_never_hash_pointers prints all addresses as unhashed

In reply to Timur Tabi
Quoted 11 lines · Timur Tabi
Quoted 5 lines · Pavel Machek

Can we make this something shorter? Clearly you don't want people placing this in their grub config, so they'll be most likely typing this a lot...

debug_pointers or debug_ptrs would be better.

dbg_unhash_ptrs? "debug_ptrs" is too vague IMHO, and I want to keep the word "hash" somewhere there to indicate exactly what's happening.

I understand that the long name is painful. But I would prefer to avoid another bike shedding over it.

There was some pushback against this feature in general. It should be used deliberately and people must be aware of the consequences. This is why it is only boot option and why it prints such a huge warning. The long clear name helps as well.

I propose to keep the name as is for now. We could always introduce an alias later when there is a wide preference and consensus.

Best Regards,
Petr

author
#11 lore raw

Re: [PATCH 3/3] [v3] lib/vsprintf: debug_never_hash_pointers prints all addresses as unhashed

Quoted 5 lines · Petr Mladek

There was some pushback against this feature in general. It should be used deliberately and people must be aware of the consequences. This is why it is only boot option and why it prints such a huge warning. The long clear name helps as well.

This is my thinking as well. I wanted it to be a bit obnoxious.

maintainer
#12 lore raw

Re: [PATCH 3/3] [v3] lib/vsprintf: debug_never_hash_pointers prints all addresses as unhashed

In reply to Timur Tabi
Quoted 35 lines · Timur Tabi

If the debug_never_hash_pointers command line parameter is set, then printk("%p") will print pointers as unhashed, which is useful for debugging purposes. This also applies to any function that uses vsprintf, such as print_hex_dump() and seq_buf_printf().

A large warning message is displayed if this option is enabled. Unhashed pointers expose kernel addresses, which can be a security risk.

Also update test_printf to skip the hashed pointer tests if the command-line option is set.

---
 .../admin-guide/kernel-parameters.txt         | 15 ++++++++
 lib/test_printf.c                             |  8 ++++
 lib/vsprintf.c                                | 38 ++++++++++++++++++-
 3 files changed, 59 insertions(+), 2 deletions(-)
Documentation/admin-guide/kernel-parameters.txt +4 -0
index a10b545c2070..2a97e787f49c 100644
@@ -810,6 +810,21 @@
 			1 will print _a lot_ more information - normally
 			only useful to kernel developers.
 
+	debug_never_hash_pointers
+			Force pointers printed to the console or buffers to be
+			unhashed.  By default, when a pointer is printed via %p
+			format string, that pointer is "hashed", i.e. obscured

Documentation/admin-guide/kernel-parameters.txt line 826

Quoted 10 lines
Documentation/admin-guide/kernel-parameters.txt +10 -0
+			by hashing the pointer value.  This is a security feature
+			that hides actual kernel addresses from unprivileged
+			users, but it also makes debugging the kernel more
+			difficult since unequal pointers can no longer be
+			compared.  However, if this command-line option is
+			specified, then all normal pointers will have their true
+			value printed.  Pointers printed via %pK may still be
+			hashed.  This option should only be specified when
+			debugging the kernel.  Please do not use on production
+			kernels.

I like this description.

Quoted 11 lines
lib/vsprintf.c +3 -0
index 3b53c73580c5..b4e07ecb1cb2 100644
@@ -2090,6 +2090,34 @@ char *fwnode_string(char *buf, char *end, struct fwnode_handle *fwnode,
 	return widen_string(buf, buf - buf_start, end, spec);
 }
 
+/* Disable pointer hashing if requested */
+bool debug_never_hash_pointers __ro_after_init;
+EXPORT_SYMBOL_GPL(debug_never_hash_pointers);

lib/vsprintf.c char *fwnode_string(char *buf, char *end, struct fwnode_handle *fwnode, · line 2105

Quoted 10 lines
lib/vsprintf.c +10 -0
+
+static int __init debug_never_hash_pointers_enable(char *str)
+{
+	debug_never_hash_pointers = true;
+
+	pr_warn("**********************************************************\n");
+	pr_warn("**   NOTICE NOTICE NOTICE NOTICE NOTICE NOTICE NOTICE   **\n");
+	pr_warn("**                                                      **\n");
+	pr_warn("** All pointers that are printed to the console will    **\n");
+	pr_warn("** be printed as unhashed.                              **\n");

I would really like to make it clear here that it is not only about consoles. Most people will see only this message. Only few people read documentation. Many people will learn the parameter name from another context by googling.

I know that it is not easy to find good words. Especially because pointers printed by %pK might still be hashed.

lib/vsprintf.c char *fwnode_string(char *buf, char *end, struct fwnode_handle *fwnode, · line 2108

Quoted 3 lines
lib/vsprintf.c +3 -0
+	pr_warn("**                                                      **\n");
+	pr_warn("** Kernel memory addresses are exposed, which may       **\n");
+	pr_warn("** reduce the security of your system.                  **\n");

What about replacing the first two paragraphs with something like:

"This system shows unhashed kernel memory addresses via logs and
other interfaces. It might reduce the security of your system."

Best Regards,
Petr

Quoted 34 lines
lib/vsprintf.c +20 -2
+	pr_warn("**                                                      **\n");
+	pr_warn("** If you see this message and you are not debugging    **\n");
+	pr_warn("** the kernel, report this immediately to your system   **\n");
+	pr_warn("** administrator!                                       **\n");
+	pr_warn("**                                                      **\n");
+	pr_warn("**   NOTICE NOTICE NOTICE NOTICE NOTICE NOTICE NOTICE   **\n");
+	pr_warn("**********************************************************\n");
+
+	return 0;
+}
+early_param("debug_never_hash_pointers", debug_never_hash_pointers_enable);
+
 /*
  * Show a '%p' thing.  A kernel extension is that the '%p' is followed
  * by an extra set of alphanumeric characters that are extended format
@@ -2297,8 +2325,14 @@ char *pointer(const char *fmt, char *buf, char *end, void *ptr,
 		}
 	}
 
-	/* default is to _not_ leak addresses, hash before printing */
-	return ptr_to_id(buf, end, ptr, spec);
+	/*
+	 * default is to _not_ leak addresses, so hash before printing,
+	 * unless debug_never_hash_pointers is specified on the command line.
+	 */
+	if (unlikely(debug_never_hash_pointers))
+		return pointer_string(buf, end, ptr, spec);
+	else
+		return ptr_to_id(buf, end, ptr, spec);
 }
 
 /*
Signature · 1 line
2.25.1
author
#13 lore raw

Re: [PATCH 3/3] [v3] lib/vsprintf: debug_never_hash_pointers prints all addresses as unhashed

Quoted 8 lines · Petr Mladek

I would really like to make it clear here that it is not only about consoles. Most people will see only this message. Only few people read documentation. Many people will learn the parameter name from another context by googling.

I know that it is not easy to find good words. Especially because pointers printed by %pK might still be hashed.

Quoted 8 lines

lib/vsprintf.c char *fwnode_string(char *buf, char *end, struct fwnode_handle *fwnode, · line 2108

Quoted 3 lines
lib/vsprintf.c +3 -0
+	pr_warn("**                                                      **\n");
+	pr_warn("** Kernel memory addresses are exposed, which may       **\n");
+	pr_warn("** reduce the security of your system.                  **\n");

What about replacing the first two paragraphs with something like:

"This system shows unhashed kernel memory addresses via logs and
other interfaces. It might reduce the security of your system."

That works, thanks. I'll post a v4 soon.

#14 lore raw
In reply to Timur Tabi
Quoted 13 lines · Timur Tabi

[The list of email addresses on CC: is getting quite lengthy, so I hope I've included everyone.]

Although hashing addresses printed via printk does make the kernel more secure, it interferes with debugging, especially with some functions like print_hex_dump() which always uses hashed addresses.

To avoid having to choose between %p and %px, it's easier to add a kernel command line that treats all %p as %px. This encourages developers to use %p more without making debugging more difficult.

Quoted 7 lines

Patches #1 and #2 upgrade the kselftest framework so that it can report on tests that were skipped outright. This is needed for the test_printf module which will now skip %p hashing tests if hashing is disabled.

Patch #2 upgrades the printf library to check the command line. It also updates test_printf().

Side note to the future contributions.

Quoted 3 lines

Full series:

That's what you usually add to each patch in the series individually. You may use simple oneliner for this, i.e. git filter-branch --msg-filter ...

Signature · 2 lines
With Best Regards,
Andy Shevchenko