linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* [PATCH v2 0/3] floppy: suppress UBSAN warning in setup_rw_floppy()
@ 2020-04-26 13:07 Denis Efremov
  2020-04-26 13:07 ` [PATCH v2 1/3] floppy: add FD_AUTODETECT_SIZE define for struct floppy_drive_params Denis Efremov
                   ` (2 more replies)
  0 siblings, 3 replies; 11+ messages in thread
From: Denis Efremov @ 2020-04-26 13:07 UTC (permalink / raw)
  To: linux-block; +Cc: Denis Efremov, Willy Tarreau, Christoph Hellwig, linux-kernel

These patches are based on Willy's cleanup patches
https://lkml.org/lkml/2020/3/31/609.

The first two patches indroduce defines to make it more clear why
cmd_count in struct floppy_raw_cmd allows out-of-bounds access for
cmd, reply_count, reply fields. The last one handles the warning.

Changes in v2:
 - add FD_RAW_CMD_FULLSIZE to the second patch
 - add union with fullcmd to struct floppy_raw_cmd in the last patch

Denis Efremov (3):
  floppy: add FD_AUTODETECT_SIZE define for struct floppy_drive_params
  floppy: add defines for sizes of cmd & reply buffers of floppy_raw_cmd
  floppy: suppress UBSAN warning in setup_rw_floppy()

 drivers/block/floppy.c  | 34 +++++++++++++---------------------
 include/uapi/linux/fd.h | 26 ++++++++++++++++++++++----
 2 files changed, 35 insertions(+), 25 deletions(-)

-- 
2.25.3


^ permalink raw reply	[flat|nested] 11+ messages in thread

* [PATCH v2 1/3] floppy: add FD_AUTODETECT_SIZE define for struct floppy_drive_params
  2020-04-26 13:07 [PATCH v2 0/3] floppy: suppress UBSAN warning in setup_rw_floppy() Denis Efremov
@ 2020-04-26 13:07 ` Denis Efremov
  2020-04-26 20:24   ` Joe Perches
  2020-04-27  5:48   ` Christoph Hellwig
  2020-04-26 13:07 ` [PATCH v2 2/3] floppy: add defines for sizes of cmd & reply buffers of floppy_raw_cmd Denis Efremov
  2020-04-26 13:07 ` [PATCH v2 3/3] floppy: suppress UBSAN warning in setup_rw_floppy() Denis Efremov
  2 siblings, 2 replies; 11+ messages in thread
From: Denis Efremov @ 2020-04-26 13:07 UTC (permalink / raw)
  To: linux-block; +Cc: Denis Efremov, Willy Tarreau, Christoph Hellwig, linux-kernel

Use FD_AUTODETECT_SIZE for autodetect buffer size in struct
floppy_drive_params instead of a magic number.

Signed-off-by: Denis Efremov <efremov@linux.com>
---
 drivers/block/floppy.c  | 9 +++++----
 include/uapi/linux/fd.h | 5 ++++-
 2 files changed, 9 insertions(+), 5 deletions(-)

diff --git a/drivers/block/floppy.c b/drivers/block/floppy.c
index 2817170dd403..ac2023c757e3 100644
--- a/drivers/block/floppy.c
+++ b/drivers/block/floppy.c
@@ -2076,7 +2076,8 @@ static int next_valid_format(int drive)
 
 	probed_format = drive_state[drive].probed_format;
 	while (1) {
-		if (probed_format >= 8 || !drive_params[drive].autodetect[probed_format]) {
+		if (probed_format >= FD_AUTODETECT_SIZE ||
+			!drive_params[drive].autodetect[probed_format]) {
 			drive_state[drive].probed_format = 0;
 			return 1;
 		}
@@ -3445,13 +3446,13 @@ static int fd_getgeo(struct block_device *bdev, struct hd_geometry *geo)
 	return 0;
 }
 
-static bool valid_floppy_drive_params(const short autodetect[8],
+static bool valid_floppy_drive_params(const short autodetect[FD_AUTODETECT_SIZE],
 		int native_format)
 {
 	size_t floppy_type_size = ARRAY_SIZE(floppy_type);
 	size_t i = 0;
 
-	for (i = 0; i < 8; ++i) {
+	for (i = 0; i < FD_AUTODETECT_SIZE; ++i) {
 		if (autodetect[i] < 0 ||
 		    autodetect[i] >= floppy_type_size)
 			return false;
@@ -3676,7 +3677,7 @@ struct compat_floppy_drive_params {
 	struct floppy_max_errors max_errors;
 	char		flags;
 	char		read_track;
-	short		autodetect[8];
+	short		autodetect[FD_AUTODETECT_SIZE];
 	compat_int_t	checkfreq;
 	compat_int_t	native_format;
 };
diff --git a/include/uapi/linux/fd.h b/include/uapi/linux/fd.h
index 90fb94712c41..3f6b7be4c096 100644
--- a/include/uapi/linux/fd.h
+++ b/include/uapi/linux/fd.h
@@ -172,7 +172,10 @@ struct floppy_drive_params {
  * used in succession to try to read the disk. If the FDC cannot lock onto
  * the disk, the next format is tried. This uses the variable 'probing'.
  */
-	short autodetect[8];		/* autodetected formats */
+
+#define FD_AUTODETECT_SIZE 8
+
+	short autodetect[FD_AUTODETECT_SIZE]; /* autodetected formats */
 	
 	int checkfreq; /* how often should the drive be checked for disk 
 			* changes */
-- 
2.25.3


^ permalink raw reply related	[flat|nested] 11+ messages in thread

* [PATCH v2 2/3] floppy: add defines for sizes of cmd & reply buffers of floppy_raw_cmd
  2020-04-26 13:07 [PATCH v2 0/3] floppy: suppress UBSAN warning in setup_rw_floppy() Denis Efremov
  2020-04-26 13:07 ` [PATCH v2 1/3] floppy: add FD_AUTODETECT_SIZE define for struct floppy_drive_params Denis Efremov
@ 2020-04-26 13:07 ` Denis Efremov
  2020-04-26 20:28   ` Joe Perches
  2020-04-27  5:50   ` Christoph Hellwig
  2020-04-26 13:07 ` [PATCH v2 3/3] floppy: suppress UBSAN warning in setup_rw_floppy() Denis Efremov
  2 siblings, 2 replies; 11+ messages in thread
From: Denis Efremov @ 2020-04-26 13:07 UTC (permalink / raw)
  To: linux-block; +Cc: Denis Efremov, Willy Tarreau, Christoph Hellwig, linux-kernel

Use FD_RAW_CMD_SIZE, FD_RAW_REPLY_SIZE defines instead of magic numbers
for cmd & reply buffers of struct floppy_raw_cmd. Remove local to
floppy.c MAX_REPLIES define, as it is now FD_RAW_REPLY_SIZE.
FD_RAW_CMD_FULLSIZE added as we allow command to also fill reply_count
and reply fields.

Signed-off-by: Denis Efremov <efremov@linux.com>
---
 drivers/block/floppy.c  | 21 ++++++---------------
 include/uapi/linux/fd.h | 14 ++++++++++++--
 2 files changed, 18 insertions(+), 17 deletions(-)

diff --git a/drivers/block/floppy.c b/drivers/block/floppy.c
index ac2023c757e3..052ba457956e 100644
--- a/drivers/block/floppy.c
+++ b/drivers/block/floppy.c
@@ -337,8 +337,7 @@ static bool initialized;
 /*
  * globals used by 'result()'
  */
-#define MAX_REPLIES 16
-static unsigned char reply_buffer[MAX_REPLIES];
+static unsigned char reply_buffer[FD_RAW_REPLY_SIZE];
 static int inr;		/* size of reply buffer, when called from interrupt */
 #define ST0		0
 #define ST1		1
@@ -1165,7 +1164,7 @@ static int result(int fdc)
 	int i;
 	int status = 0;
 
-	for (i = 0; i < MAX_REPLIES; i++) {
+	for (i = 0; i < FD_RAW_REPLY_SIZE; i++) {
 		status = wait_til_ready(fdc);
 		if (status < 0)
 			break;
@@ -1847,7 +1846,7 @@ static void show_floppy(int fdc)
 			output_log[(i + output_log_pos) % OLOGSIZE].jiffies);
 	pr_info("last result at %lu\n", resultjiffies);
 	pr_info("last redo_fd_request at %lu\n", lastredo);
-	print_hex_dump(KERN_INFO, "", DUMP_PREFIX_NONE, 16, 1,
+	print_hex_dump(KERN_INFO, "", DUMP_PREFIX_NONE, FD_RAW_REPLY_SIZE, 1,
 		       reply_buffer, resultsize, true);
 
 	pr_info("status=%x\n", fdc_inb(fdc, FD_STATUS));
@@ -3082,7 +3081,7 @@ static void raw_cmd_done(int flag)
 		raw_cmd->flags |= FD_RAW_HARDFAILURE;
 	} else {
 		raw_cmd->reply_count = inr;
-		if (raw_cmd->reply_count > MAX_REPLIES)
+		if (raw_cmd->reply_count > FD_RAW_REPLY_SIZE)
 			raw_cmd->reply_count = 0;
 		for (i = 0; i < raw_cmd->reply_count; i++)
 			raw_cmd->reply[i] = reply_buffer[i];
@@ -3193,18 +3192,10 @@ static int raw_cmd_copyin(int cmd, void __user *param,
 	if (ret)
 		return -EFAULT;
 	param += sizeof(struct floppy_raw_cmd);
-	if (ptr->cmd_count > 33)
-			/* the command may now also take up the space
-			 * initially intended for the reply & the
-			 * reply count. Needed for long 82078 commands
-			 * such as RESTORE, which takes ... 17 command
-			 * bytes. Murphy's law #137: When you reserve
-			 * 16 bytes for a structure, you'll one day
-			 * discover that you really need 17...
-			 */
+	if (ptr->cmd_count > FD_RAW_CMD_FULLSIZE)
 		return -EINVAL;
 
-	for (i = 0; i < 16; i++)
+	for (i = 0; i < FD_RAW_REPLY_SIZE; i++)
 		ptr->reply[i] = 0;
 	ptr->resultcode = 0;
 
diff --git a/include/uapi/linux/fd.h b/include/uapi/linux/fd.h
index 3f6b7be4c096..2e9c2c1c18e6 100644
--- a/include/uapi/linux/fd.h
+++ b/include/uapi/linux/fd.h
@@ -360,10 +360,20 @@ struct floppy_raw_cmd {
 	int buffer_length; /* length of allocated buffer */
 
 	unsigned char rate;
+
+#define FD_RAW_CMD_SIZE 16
+#define FD_RAW_REPLY_SIZE 16
+#define FD_RAW_CMD_FULLSIZE (FD_RAW_CMD_SIZE + 1 + FD_RAW_REPLY_SIZE)
+
+	/* The command may take up the space initially intended for the reply
+	 * and the reply count. Needed for long 82078 commands such as RESTORE,
+	 * which takes 17 command bytes.
+	 */
+
 	unsigned char cmd_count;
-	unsigned char cmd[16];
+	unsigned char cmd[FD_RAW_CMD_SIZE];
 	unsigned char reply_count;
-	unsigned char reply[16];
+	unsigned char reply[FD_RAW_REPLY_SIZE];
 	int track;
 	int resultcode;
 
-- 
2.25.3


^ permalink raw reply related	[flat|nested] 11+ messages in thread

* [PATCH v2 3/3] floppy: suppress UBSAN warning in setup_rw_floppy()
  2020-04-26 13:07 [PATCH v2 0/3] floppy: suppress UBSAN warning in setup_rw_floppy() Denis Efremov
  2020-04-26 13:07 ` [PATCH v2 1/3] floppy: add FD_AUTODETECT_SIZE define for struct floppy_drive_params Denis Efremov
  2020-04-26 13:07 ` [PATCH v2 2/3] floppy: add defines for sizes of cmd & reply buffers of floppy_raw_cmd Denis Efremov
@ 2020-04-26 13:07 ` Denis Efremov
  2020-04-26 23:19   ` Joe Perches
  2020-04-27  5:50   ` Christoph Hellwig
  2 siblings, 2 replies; 11+ messages in thread
From: Denis Efremov @ 2020-04-26 13:07 UTC (permalink / raw)
  To: linux-block; +Cc: Denis Efremov, Willy Tarreau, Christoph Hellwig, linux-kernel

UBSAN: array-index-out-of-bounds in drivers/block/floppy.c:1521:45
index 16 is out of range for type 'unsigned char [16]'
Call Trace:
...
 setup_rw_floppy+0x5c3/0x7f0
 floppy_ready+0x2be/0x13b0
 process_one_work+0x2c1/0x5d0
 worker_thread+0x56/0x5e0
 kthread+0x122/0x170
 ret_from_fork+0x35/0x40

From include/uapi/linux/fd.h:
struct floppy_raw_cmd {
	...
	unsigned char cmd_count;
	unsigned char cmd[16];
	unsigned char reply_count;
	unsigned char reply[16];
	...
}

This out-of-bounds access is intentional. The command in struct
floppy_raw_cmd may take up the space initially intended for the reply
and the reply count. It is needed for long 82078 commands such as
RESTORE, which takes 17 command bytes. Initial cmd size is not enough
and since struct setup_rw_floppy is a part of uapi we check that
cmd_count is in [0:16+1+16] in raw_cmd_copyin().

The patch adds union with original cmd,reply_count,reply fields and
fullcmd field of equivalent size. The cmd accesses are turned to
fullcmd where appropriate to suppress UBSAN warning.

Signed-off-by: Denis Efremov <efremov@linux.com>
---
 drivers/block/floppy.c  |  4 ++--
 include/uapi/linux/fd.h | 11 ++++++++---
 2 files changed, 10 insertions(+), 5 deletions(-)

diff --git a/drivers/block/floppy.c b/drivers/block/floppy.c
index 052ba457956e..1b22aab6cd9b 100644
--- a/drivers/block/floppy.c
+++ b/drivers/block/floppy.c
@@ -1072,7 +1072,7 @@ static void setup_DMA(void)
 
 		pr_info("zero dma transfer size:");
 		for (i = 0; i < raw_cmd->cmd_count; i++)
-			pr_cont("%x,", raw_cmd->cmd[i]);
+			pr_cont("%x,", raw_cmd->fullcmd[i]);
 		pr_cont("\n");
 		cont->done(0);
 		fdc_state[current_fdc].reset = 1;
@@ -1518,7 +1518,7 @@ static void setup_rw_floppy(void)
 
 	r = 0;
 	for (i = 0; i < raw_cmd->cmd_count; i++)
-		r |= output_byte(current_fdc, raw_cmd->cmd[i]);
+		r |= output_byte(current_fdc, raw_cmd->fullcmd[i]);
 
 	debugt(__func__, "rw_command");
 
diff --git a/include/uapi/linux/fd.h b/include/uapi/linux/fd.h
index 2e9c2c1c18e6..8b80c63b971c 100644
--- a/include/uapi/linux/fd.h
+++ b/include/uapi/linux/fd.h
@@ -371,9 +371,14 @@ struct floppy_raw_cmd {
 	 */
 
 	unsigned char cmd_count;
-	unsigned char cmd[FD_RAW_CMD_SIZE];
-	unsigned char reply_count;
-	unsigned char reply[FD_RAW_REPLY_SIZE];
+	union {
+		struct {
+			unsigned char cmd[FD_RAW_CMD_SIZE];
+			unsigned char reply_count;
+			unsigned char reply[FD_RAW_REPLY_SIZE];
+		};
+		unsigned char fullcmd[FD_RAW_CMD_FULLSIZE];
+	};
 	int track;
 	int resultcode;
 
-- 
2.25.3


^ permalink raw reply related	[flat|nested] 11+ messages in thread

* Re: [PATCH v2 1/3] floppy: add FD_AUTODETECT_SIZE define for struct floppy_drive_params
  2020-04-26 13:07 ` [PATCH v2 1/3] floppy: add FD_AUTODETECT_SIZE define for struct floppy_drive_params Denis Efremov
@ 2020-04-26 20:24   ` Joe Perches
  2020-04-27  5:48   ` Christoph Hellwig
  1 sibling, 0 replies; 11+ messages in thread
From: Joe Perches @ 2020-04-26 20:24 UTC (permalink / raw)
  To: Denis Efremov, linux-block; +Cc: Willy Tarreau, Christoph Hellwig, linux-kernel

On Sun, 2020-04-26 at 16:07 +0300, Denis Efremov wrote:
> Use FD_AUTODETECT_SIZE for autodetect buffer size in struct
> floppy_drive_params instead of a magic number.

trivia:

> diff --git a/drivers/block/floppy.c b/drivers/block/floppy.c
[]
> @@ -2076,7 +2076,8 @@ static int next_valid_format(int drive)
>  
>  	probed_format = drive_state[drive].probed_format;
>  	while (1) {
> -		if (probed_format >= 8 || !drive_params[drive].autodetect[probed_format]) {
> +		if (probed_format >= FD_AUTODETECT_SIZE ||
> +			!drive_params[drive].autodetect[probed_format]) {

floppy.c almost exclusively uses multi-line statement
alignment to open parenthesis.  It'd be nicer to keep
using that style.


>  			drive_state[drive].probed_format = 0;
>  			return 1;
>  		}
> @@ -3445,13 +3446,13 @@ static int fd_getgeo(struct block_device *bdev, struct hd_geometry *geo)
>  	return 0;
>  }
>  
> -static bool valid_floppy_drive_params(const short autodetect[8],
> +static bool valid_floppy_drive_params(const short autodetect[FD_AUTODETECT_SIZE],
>  		int native_format)
> 


^ permalink raw reply	[flat|nested] 11+ messages in thread

* Re: [PATCH v2 2/3] floppy: add defines for sizes of cmd & reply buffers of floppy_raw_cmd
  2020-04-26 13:07 ` [PATCH v2 2/3] floppy: add defines for sizes of cmd & reply buffers of floppy_raw_cmd Denis Efremov
@ 2020-04-26 20:28   ` Joe Perches
  2020-04-26 20:44     ` Denis Efremov
  2020-04-27  5:50   ` Christoph Hellwig
  1 sibling, 1 reply; 11+ messages in thread
From: Joe Perches @ 2020-04-26 20:28 UTC (permalink / raw)
  To: Denis Efremov, linux-block; +Cc: Willy Tarreau, Christoph Hellwig, linux-kernel

On Sun, 2020-04-26 at 16:07 +0300, Denis Efremov wrote:
> Use FD_RAW_CMD_SIZE, FD_RAW_REPLY_SIZE defines instead of magic numbers
> for cmd & reply buffers of struct floppy_raw_cmd. Remove local to
> floppy.c MAX_REPLIES define, as it is now FD_RAW_REPLY_SIZE.
> FD_RAW_CMD_FULLSIZE added as we allow command to also fill reply_count
> and reply fields.
[]
> diff --git a/drivers/block/floppy.c b/drivers/block/floppy.c
[]
> @@ -1847,7 +1846,7 @@ static void show_floppy(int fdc)
>  			output_log[(i + output_log_pos) % OLOGSIZE].jiffies);
>  	pr_info("last result at %lu\n", resultjiffies);
>  	pr_info("last redo_fd_request at %lu\n", lastredo);
> -	print_hex_dump(KERN_INFO, "", DUMP_PREFIX_NONE, 16, 1,
> +	print_hex_dump(KERN_INFO, "", DUMP_PREFIX_NONE, FD_RAW_REPLY_SIZE, 1,
>  		       reply_buffer, resultsize, true);

FD_RAW_REPLY_SIZE happens to be 16, but it's misleading
to use it here.

This use of 16 is not for FD_RAW_REPLY_SIZE, but the
width of the line
being dumped, and this value must be
either 16 or 32.



^ permalink raw reply	[flat|nested] 11+ messages in thread

* Re: [PATCH v2 2/3] floppy: add defines for sizes of cmd & reply buffers of floppy_raw_cmd
  2020-04-26 20:28   ` Joe Perches
@ 2020-04-26 20:44     ` Denis Efremov
  0 siblings, 0 replies; 11+ messages in thread
From: Denis Efremov @ 2020-04-26 20:44 UTC (permalink / raw)
  To: Joe Perches, linux-block; +Cc: Willy Tarreau, Christoph Hellwig, linux-kernel



On 4/26/20 11:28 PM, Joe Perches wrote:
> On Sun, 2020-04-26 at 16:07 +0300, Denis Efremov wrote:
>> Use FD_RAW_CMD_SIZE, FD_RAW_REPLY_SIZE defines instead of magic numbers
>> for cmd & reply buffers of struct floppy_raw_cmd. Remove local to
>> floppy.c MAX_REPLIES define, as it is now FD_RAW_REPLY_SIZE.
>> FD_RAW_CMD_FULLSIZE added as we allow command to also fill reply_count
>> and reply fields.
> []
>> diff --git a/drivers/block/floppy.c b/drivers/block/floppy.c
> []
>> @@ -1847,7 +1846,7 @@ static void show_floppy(int fdc)
>>  			output_log[(i + output_log_pos) % OLOGSIZE].jiffies);
>>  	pr_info("last result at %lu\n", resultjiffies);
>>  	pr_info("last redo_fd_request at %lu\n", lastredo);
>> -	print_hex_dump(KERN_INFO, "", DUMP_PREFIX_NONE, 16, 1,
>> +	print_hex_dump(KERN_INFO, "", DUMP_PREFIX_NONE, FD_RAW_REPLY_SIZE, 1,
>>  		       reply_buffer, resultsize, true);
> 
> FD_RAW_REPLY_SIZE happens to be 16, but it's misleading
> to use it here.
> 
> This use of 16 is not for FD_RAW_REPLY_SIZE, but the
> width of the line
> being dumped, and this value must be
> either 16 or 32.
> 

Yes, you are right. Thanks for catching. I will resend the patches.

Denis
 

^ permalink raw reply	[flat|nested] 11+ messages in thread

* Re: [PATCH v2 3/3] floppy: suppress UBSAN warning in setup_rw_floppy()
  2020-04-26 13:07 ` [PATCH v2 3/3] floppy: suppress UBSAN warning in setup_rw_floppy() Denis Efremov
@ 2020-04-26 23:19   ` Joe Perches
  2020-04-27  5:50   ` Christoph Hellwig
  1 sibling, 0 replies; 11+ messages in thread
From: Joe Perches @ 2020-04-26 23:19 UTC (permalink / raw)
  To: Denis Efremov, linux-block; +Cc: Willy Tarreau, Christoph Hellwig, linux-kernel

On Sun, 2020-04-26 at 16:07 +0300, Denis Efremov wrote:
> UBSAN: array-index-out-of-bounds in drivers/block/floppy.c:1521:45
> index 16 is out of range for type 'unsigned char [16]'
[]
> This out-of-bounds access is intentional. The command in struct
> floppy_raw_cmd may take up the space initially intended for the reply
> and the reply count. It is needed for long 82078 commands such as
> RESTORE, which takes 17 command bytes. Initial cmd size is not enough
> and since struct setup_rw_floppy is a part of uapi we check that
> cmd_count is in [0:16+1+16] in raw_cmd_copyin().
> 
> The patch adds union with original cmd,reply_count,reply fields and
> fullcmd field of equivalent size. The cmd accesses are turned to
> fullcmd where appropriate to suppress UBSAN warning.
[]
> diff --git a/drivers/block/floppy.c b/drivers/block/floppy.c
[]
> @@ -1072,7 +1072,7 @@ static void setup_DMA(void)
>  
>  		pr_info("zero dma transfer size:");
>  		for (i = 0; i < raw_cmd->cmd_count; i++)
> -			pr_cont("%x,", raw_cmd->cmd[i]);
> +			pr_cont("%x,", raw_cmd->fullcmd[i]);
>  		pr_cont("\n");

slightly unrelated trivia: perhaps better as:

		print_hex_dump(KERN_INFO, "zero dma transfer size: ",
			       DUMP_PREFIX_NONE, 16, 1,
			       raw_cmd->fullcmd, raw_cmd->cmd_count, false);

to avoid pr_cont use.



^ permalink raw reply	[flat|nested] 11+ messages in thread

* Re: [PATCH v2 1/3] floppy: add FD_AUTODETECT_SIZE define for struct floppy_drive_params
  2020-04-26 13:07 ` [PATCH v2 1/3] floppy: add FD_AUTODETECT_SIZE define for struct floppy_drive_params Denis Efremov
  2020-04-26 20:24   ` Joe Perches
@ 2020-04-27  5:48   ` Christoph Hellwig
  1 sibling, 0 replies; 11+ messages in thread
From: Christoph Hellwig @ 2020-04-27  5:48 UTC (permalink / raw)
  To: Denis Efremov; +Cc: linux-block, Willy Tarreau, Christoph Hellwig, linux-kernel

On Sun, Apr 26, 2020 at 04:07:26PM +0300, Denis Efremov wrote:
> Use FD_AUTODETECT_SIZE for autodetect buffer size in struct
> floppy_drive_params instead of a magic number.
> 
> Signed-off-by: Denis Efremov <efremov@linux.com>
> ---
>  drivers/block/floppy.c  | 9 +++++----
>  include/uapi/linux/fd.h | 5 ++++-
>  2 files changed, 9 insertions(+), 5 deletions(-)
> 
> diff --git a/drivers/block/floppy.c b/drivers/block/floppy.c
> index 2817170dd403..ac2023c757e3 100644
> --- a/drivers/block/floppy.c
> +++ b/drivers/block/floppy.c
> @@ -2076,7 +2076,8 @@ static int next_valid_format(int drive)
>  
>  	probed_format = drive_state[drive].probed_format;
>  	while (1) {
> -		if (probed_format >= 8 || !drive_params[drive].autodetect[probed_format]) {
> +		if (probed_format >= FD_AUTODETECT_SIZE ||
> +			!drive_params[drive].autodetect[probed_format]) {
>  			drive_state[drive].probed_format = 0;
>  			return 1;
>  		}
> @@ -3445,13 +3446,13 @@ static int fd_getgeo(struct block_device *bdev, struct hd_geometry *geo)
>  	return 0;
>  }
>  
> -static bool valid_floppy_drive_params(const short autodetect[8],
> +static bool valid_floppy_drive_params(const short autodetect[FD_AUTODETECT_SIZE],

This adds an > 80 char line.  But then again passing array sizes to
functions is a bit pointless ayway.

Otherwise this looks fine to me.

^ permalink raw reply	[flat|nested] 11+ messages in thread

* Re: [PATCH v2 2/3] floppy: add defines for sizes of cmd & reply buffers of floppy_raw_cmd
  2020-04-26 13:07 ` [PATCH v2 2/3] floppy: add defines for sizes of cmd & reply buffers of floppy_raw_cmd Denis Efremov
  2020-04-26 20:28   ` Joe Perches
@ 2020-04-27  5:50   ` Christoph Hellwig
  1 sibling, 0 replies; 11+ messages in thread
From: Christoph Hellwig @ 2020-04-27  5:50 UTC (permalink / raw)
  To: Denis Efremov; +Cc: linux-block, Willy Tarreau, Christoph Hellwig, linux-kernel

Looks good modulo the hexdump nitpick:

Reviewed-by: Christoph Hellwig <hch@lst.de>

^ permalink raw reply	[flat|nested] 11+ messages in thread

* Re: [PATCH v2 3/3] floppy: suppress UBSAN warning in setup_rw_floppy()
  2020-04-26 13:07 ` [PATCH v2 3/3] floppy: suppress UBSAN warning in setup_rw_floppy() Denis Efremov
  2020-04-26 23:19   ` Joe Perches
@ 2020-04-27  5:50   ` Christoph Hellwig
  1 sibling, 0 replies; 11+ messages in thread
From: Christoph Hellwig @ 2020-04-27  5:50 UTC (permalink / raw)
  To: Denis Efremov; +Cc: linux-block, Willy Tarreau, Christoph Hellwig, linux-kernel

Looks good,

Reviewed-by: Christoph Hellwig <hch@lst.de>

^ permalink raw reply	[flat|nested] 11+ messages in thread

end of thread, other threads:[~2020-04-27  5:50 UTC | newest]

Thread overview: 11+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2020-04-26 13:07 [PATCH v2 0/3] floppy: suppress UBSAN warning in setup_rw_floppy() Denis Efremov
2020-04-26 13:07 ` [PATCH v2 1/3] floppy: add FD_AUTODETECT_SIZE define for struct floppy_drive_params Denis Efremov
2020-04-26 20:24   ` Joe Perches
2020-04-27  5:48   ` Christoph Hellwig
2020-04-26 13:07 ` [PATCH v2 2/3] floppy: add defines for sizes of cmd & reply buffers of floppy_raw_cmd Denis Efremov
2020-04-26 20:28   ` Joe Perches
2020-04-26 20:44     ` Denis Efremov
2020-04-27  5:50   ` Christoph Hellwig
2020-04-26 13:07 ` [PATCH v2 3/3] floppy: suppress UBSAN warning in setup_rw_floppy() Denis Efremov
2020-04-26 23:19   ` Joe Perches
2020-04-27  5:50   ` Christoph Hellwig

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).