Didn't notice that. I agree that this will result only into redundancy. Quick look over files reveal that there are multiple places where people are using print statements after memory allocation fails. Should I go ahead and send patches to remove those redundant print statements.

On Wed, 10 Jul 2019 at 12:35, Greg Kroah-Hartman gregkh@linuxfoundation.org wrote:

On Wed, Jul 10, 2019 at 08:20:17AM -0400, Keyur Patel wrote:

...

Added missing logging statement when kfifo_alloc fails, to improve debugging.

Signed-off-by: Keyur Patel iamkeyur96@gmail.com

drivers/staging/greybus/uart.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/drivers/staging/greybus/uart.c

b/drivers/staging/greybus/uart.c

...

index b3bffe91ae99..86a395ae177d 100644 --- a/drivers/staging/greybus/uart.c +++ b/drivers/staging/greybus/uart.c @@ -856,8 +856,10 @@ static int gb_uart_probe(struct gbphy_device

*gbphy_dev,

...

  retval = kfifo_alloc(&gb_tty->write_fifo, GB_UART_WRITE_FIFO_SIZE,
                       GFP_KERNEL);

• if (retval)
  

• if (retval) {
  

•         pr_err("kfifo_alloc failed\n");
          goto exit_buf_free;
  

• }
  

You should have already gotten an error message from the log if this fails, from the kmalloc_array() call failing, right?

So why is this needed? We have been trying to remove these types of messages and keep them in the "root" place where the failure happens.

thanks,

greg k-h

On Wed, Jul 10, 2019 at 10:24:06AM -0400, Keyur Patel wrote:

On Wed, Jul 10, 2019 at 06:35:38PM +0200, Greg Kroah-Hartman wrote:

...

On Wed, Jul 10, 2019 at 08:20:17AM -0400, Keyur Patel wrote:

...

Added missing logging statement when kfifo_alloc fails, to improve debugging.

Signed-off-by: Keyur Patel iamkeyur96@gmail.com

drivers/staging/greybus/uart.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/drivers/staging/greybus/uart.c b/drivers/staging/greybus/uart.c index b3bffe91ae99..86a395ae177d 100644 --- a/drivers/staging/greybus/uart.c +++ b/drivers/staging/greybus/uart.c @@ -856,8 +856,10 @@ static int gb_uart_probe(struct gbphy_device *gbphy_dev, retval = kfifo_alloc(&gb_tty->write_fifo, GB_UART_WRITE_FIFO_SIZE, GFP_KERNEL);

• if (retval)

• if (retval) {

• pr_err("kfifo_alloc failed\n");
  

  goto exit_buf_free;
• }

You should have already gotten an error message from the log if this fails, from the kmalloc_array() call failing, right?

So why is this needed? We have been trying to remove these types of messages and keep them in the "root" place where the failure happens.

thanks,

greg k-h

Didn't notice that. I agree that this will result only into redundancy. Quick look over files reveal that there are multiple places where people are using print statements after memory allocation fails. Should I go ahead and send patches to remove those redundant print statements?

It all depends on where in the kernel those are. Some subsystems want minor cleanups like this, but most do not. So unless the issue you find is in the drivers/staging/ directory, I would not worry about it.

thanks,

greg k-h