* [PATCH 2/5] tty: srmcons: make srmcons_do_write() return void
[not found] <20231127123713.14504-1-jirislaby@kernel.org>
@ 2023-11-27 12:37 ` Jiri Slaby (SUSE)
2023-11-28 13:25 ` Ilpo Järvinen
2023-11-27 12:37 ` [PATCH 3/5] tty: srmcons: use 'count' directly in srmcons_do_write() Jiri Slaby (SUSE)
` (2 subsequent siblings)
3 siblings, 1 reply; 8+ messages in thread
From: Jiri Slaby (SUSE) @ 2023-11-27 12:37 UTC (permalink / raw)
To: gregkh
Cc: linux-serial, linux-kernel, Jiri Slaby (SUSE), Richard Henderson,
Ivan Kokshaysky, Matt Turner, linux-alpha
The return value of srmcons_do_write() is ignored as all characters are
pushed. So make srmcons_do_write() to return void.
Signed-off-by: Jiri Slaby (SUSE) <jirislaby@kernel.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Cc: Ivan Kokshaysky <ink@jurassic.park.msu.ru>
Cc: Matt Turner <mattst88@gmail.com>
Cc: linux-alpha@vger.kernel.org
---
Notes:
[v2] reordered so that it makes sense
arch/alpha/kernel/srmcons.c | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
diff --git a/arch/alpha/kernel/srmcons.c b/arch/alpha/kernel/srmcons.c
index b68c5af083cd..de896fa9829e 100644
--- a/arch/alpha/kernel/srmcons.c
+++ b/arch/alpha/kernel/srmcons.c
@@ -88,7 +88,7 @@ srmcons_receive_chars(struct timer_list *t)
}
/* called with callback_lock held */
-static int
+static void
srmcons_do_write(struct tty_port *port, const char *buf, int count)
{
static char str_cr[1] = "\r";
@@ -125,7 +125,6 @@ srmcons_do_write(struct tty_port *port, const char *buf, int count)
need_cr = 0;
}
}
- return count;
}
static ssize_t
--
2.42.1
^ permalink raw reply related [flat|nested] 8+ messages in thread* Re: [PATCH 2/5] tty: srmcons: make srmcons_do_write() return void
2023-11-27 12:37 ` [PATCH 2/5] tty: srmcons: make srmcons_do_write() return void Jiri Slaby (SUSE)
@ 2023-11-28 13:25 ` Ilpo Järvinen
0 siblings, 0 replies; 8+ messages in thread
From: Ilpo Järvinen @ 2023-11-28 13:25 UTC (permalink / raw)
To: Jiri Slaby (SUSE)
Cc: Greg Kroah-Hartman, linux-serial, LKML, Richard Henderson,
Ivan Kokshaysky, Matt Turner, linux-alpha
[-- Attachment #1: Type: text/plain, Size: 1257 bytes --]
On Mon, 27 Nov 2023, Jiri Slaby (SUSE) wrote:
> The return value of srmcons_do_write() is ignored as all characters are
> pushed. So make srmcons_do_write() to return void.
>
> Signed-off-by: Jiri Slaby (SUSE) <jirislaby@kernel.org>
> Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
> Cc: Ivan Kokshaysky <ink@jurassic.park.msu.ru>
> Cc: Matt Turner <mattst88@gmail.com>
> Cc: linux-alpha@vger.kernel.org
> ---
>
> Notes:
> [v2] reordered so that it makes sense
>
> arch/alpha/kernel/srmcons.c | 3 +--
> 1 file changed, 1 insertion(+), 2 deletions(-)
>
> diff --git a/arch/alpha/kernel/srmcons.c b/arch/alpha/kernel/srmcons.c
> index b68c5af083cd..de896fa9829e 100644
> --- a/arch/alpha/kernel/srmcons.c
> +++ b/arch/alpha/kernel/srmcons.c
> @@ -88,7 +88,7 @@ srmcons_receive_chars(struct timer_list *t)
> }
>
> /* called with callback_lock held */
> -static int
> +static void
> srmcons_do_write(struct tty_port *port, const char *buf, int count)
> {
> static char str_cr[1] = "\r";
> @@ -125,7 +125,6 @@ srmcons_do_write(struct tty_port *port, const char *buf, int count)
> need_cr = 0;
> }
> }
> - return count;
> }
>
> static ssize_t
>
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
--
i.
^ permalink raw reply [flat|nested] 8+ messages in thread
* [PATCH 3/5] tty: srmcons: use 'count' directly in srmcons_do_write()
[not found] <20231127123713.14504-1-jirislaby@kernel.org>
2023-11-27 12:37 ` [PATCH 2/5] tty: srmcons: make srmcons_do_write() return void Jiri Slaby (SUSE)
@ 2023-11-27 12:37 ` Jiri Slaby (SUSE)
2023-11-28 13:26 ` Ilpo Järvinen
2023-11-27 12:37 ` [PATCH 4/5] tty: srmcons: switch need_cr to bool Jiri Slaby (SUSE)
2023-11-27 12:37 ` [PATCH 5/5] tty: srmcons: remove 'str_cr' and use string directly Jiri Slaby (SUSE)
3 siblings, 1 reply; 8+ messages in thread
From: Jiri Slaby (SUSE) @ 2023-11-27 12:37 UTC (permalink / raw)
To: gregkh
Cc: linux-serial, linux-kernel, Jiri Slaby (SUSE), Richard Henderson,
Ivan Kokshaysky, Matt Turner, linux-alpha
Similarly to 'buf' in the previous patch, there is no need to have a
separate counter ('remaining') in srmcons_do_write(). 'count' can be
used directly which simplifies the code a bit.
Note that the type of the current count ('c') is changed from 'long' to
'size_t' so that:
1) it is prepared for the upcoming change of 'count's type, and
2) is unsigned.
Signed-off-by: Jiri Slaby (SUSE) <jirislaby@kernel.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Cc: Ivan Kokshaysky <ink@jurassic.park.msu.ru>
Cc: Matt Turner <mattst88@gmail.com>
Cc: linux-alpha@vger.kernel.org
---
Notes:
[v2] reordered so that it makes sense
arch/alpha/kernel/srmcons.c | 8 ++++----
1 file changed, 4 insertions(+), 4 deletions(-)
diff --git a/arch/alpha/kernel/srmcons.c b/arch/alpha/kernel/srmcons.c
index de896fa9829e..32bc098de7da 100644
--- a/arch/alpha/kernel/srmcons.c
+++ b/arch/alpha/kernel/srmcons.c
@@ -92,24 +92,24 @@ static void
srmcons_do_write(struct tty_port *port, const char *buf, int count)
{
static char str_cr[1] = "\r";
- long c, remaining = count;
+ size_t c;
srmcons_result result;
int need_cr;
- while (remaining > 0) {
+ while (count > 0) {
need_cr = 0;
/*
* Break it up into reasonable size chunks to allow a chance
* for input to get in
*/
- for (c = 0; c < min_t(long, 128L, remaining) && !need_cr; c++)
+ for (c = 0; c < min_t(size_t, 128U, count) && !need_cr; c++)
if (buf[c] == '\n')
need_cr = 1;
while (c > 0) {
result.as_long = callback_puts(0, buf, c);
c -= result.bits.c;
- remaining -= result.bits.c;
+ count -= result.bits.c;
buf += result.bits.c;
/*
--
2.42.1
^ permalink raw reply related [flat|nested] 8+ messages in thread* Re: [PATCH 3/5] tty: srmcons: use 'count' directly in srmcons_do_write()
2023-11-27 12:37 ` [PATCH 3/5] tty: srmcons: use 'count' directly in srmcons_do_write() Jiri Slaby (SUSE)
@ 2023-11-28 13:26 ` Ilpo Järvinen
0 siblings, 0 replies; 8+ messages in thread
From: Ilpo Järvinen @ 2023-11-28 13:26 UTC (permalink / raw)
To: Jiri Slaby (SUSE)
Cc: gregkh, linux-serial, linux-kernel, Richard Henderson,
Ivan Kokshaysky, Matt Turner, linux-alpha
[-- Attachment #1: Type: text/plain, Size: 1894 bytes --]
On Mon, 27 Nov 2023, Jiri Slaby (SUSE) wrote:
> Similarly to 'buf' in the previous patch, there is no need to have a
> separate counter ('remaining') in srmcons_do_write(). 'count' can be
> used directly which simplifies the code a bit.
>
> Note that the type of the current count ('c') is changed from 'long' to
> 'size_t' so that:
> 1) it is prepared for the upcoming change of 'count's type, and
> 2) is unsigned.
>
> Signed-off-by: Jiri Slaby (SUSE) <jirislaby@kernel.org>
> Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
> Cc: Ivan Kokshaysky <ink@jurassic.park.msu.ru>
> Cc: Matt Turner <mattst88@gmail.com>
> Cc: linux-alpha@vger.kernel.org
> ---
>
> Notes:
> [v2] reordered so that it makes sense
>
> arch/alpha/kernel/srmcons.c | 8 ++++----
> 1 file changed, 4 insertions(+), 4 deletions(-)
>
> diff --git a/arch/alpha/kernel/srmcons.c b/arch/alpha/kernel/srmcons.c
> index de896fa9829e..32bc098de7da 100644
> --- a/arch/alpha/kernel/srmcons.c
> +++ b/arch/alpha/kernel/srmcons.c
> @@ -92,24 +92,24 @@ static void
> srmcons_do_write(struct tty_port *port, const char *buf, int count)
> {
> static char str_cr[1] = "\r";
> - long c, remaining = count;
> + size_t c;
> srmcons_result result;
> int need_cr;
>
> - while (remaining > 0) {
> + while (count > 0) {
> need_cr = 0;
> /*
> * Break it up into reasonable size chunks to allow a chance
> * for input to get in
> */
> - for (c = 0; c < min_t(long, 128L, remaining) && !need_cr; c++)
> + for (c = 0; c < min_t(size_t, 128U, count) && !need_cr; c++)
> if (buf[c] == '\n')
> need_cr = 1;
>
> while (c > 0) {
> result.as_long = callback_puts(0, buf, c);
> c -= result.bits.c;
> - remaining -= result.bits.c;
> + count -= result.bits.c;
> buf += result.bits.c;
>
> /*
>
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
--
i.
^ permalink raw reply [flat|nested] 8+ messages in thread
* [PATCH 4/5] tty: srmcons: switch need_cr to bool
[not found] <20231127123713.14504-1-jirislaby@kernel.org>
2023-11-27 12:37 ` [PATCH 2/5] tty: srmcons: make srmcons_do_write() return void Jiri Slaby (SUSE)
2023-11-27 12:37 ` [PATCH 3/5] tty: srmcons: use 'count' directly in srmcons_do_write() Jiri Slaby (SUSE)
@ 2023-11-27 12:37 ` Jiri Slaby (SUSE)
2023-11-28 13:25 ` Ilpo Järvinen
2023-11-27 12:37 ` [PATCH 5/5] tty: srmcons: remove 'str_cr' and use string directly Jiri Slaby (SUSE)
3 siblings, 1 reply; 8+ messages in thread
From: Jiri Slaby (SUSE) @ 2023-11-27 12:37 UTC (permalink / raw)
To: gregkh
Cc: linux-serial, linux-kernel, Jiri Slaby (SUSE), Richard Henderson,
Ivan Kokshaysky, Matt Turner, linux-alpha
'need_cr' is a flag, so type it properly to be a 'bool'. Move the
declaration into the loop too. That ensures the variable is initialized
properly even if the code was moved somehow.
Signed-off-by: Jiri Slaby (SUSE) <jirislaby@kernel.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Cc: Ivan Kokshaysky <ink@jurassic.park.msu.ru>
Cc: Matt Turner <mattst88@gmail.com>
Cc: linux-alpha@vger.kernel.org
---
arch/alpha/kernel/srmcons.c | 7 +++----
1 file changed, 3 insertions(+), 4 deletions(-)
diff --git a/arch/alpha/kernel/srmcons.c b/arch/alpha/kernel/srmcons.c
index 32bc098de7da..c6b821afbfd3 100644
--- a/arch/alpha/kernel/srmcons.c
+++ b/arch/alpha/kernel/srmcons.c
@@ -94,17 +94,16 @@ srmcons_do_write(struct tty_port *port, const char *buf, int count)
static char str_cr[1] = "\r";
size_t c;
srmcons_result result;
- int need_cr;
while (count > 0) {
- need_cr = 0;
+ bool need_cr = false;
/*
* Break it up into reasonable size chunks to allow a chance
* for input to get in
*/
for (c = 0; c < min_t(size_t, 128U, count) && !need_cr; c++)
if (buf[c] == '\n')
- need_cr = 1;
+ need_cr = true;
while (c > 0) {
result.as_long = callback_puts(0, buf, c);
@@ -122,7 +121,7 @@ srmcons_do_write(struct tty_port *port, const char *buf, int count)
while (need_cr) {
result.as_long = callback_puts(0, str_cr, 1);
if (result.bits.c > 0)
- need_cr = 0;
+ need_cr = false;
}
}
}
--
2.42.1
^ permalink raw reply related [flat|nested] 8+ messages in thread* Re: [PATCH 4/5] tty: srmcons: switch need_cr to bool
2023-11-27 12:37 ` [PATCH 4/5] tty: srmcons: switch need_cr to bool Jiri Slaby (SUSE)
@ 2023-11-28 13:25 ` Ilpo Järvinen
0 siblings, 0 replies; 8+ messages in thread
From: Ilpo Järvinen @ 2023-11-28 13:25 UTC (permalink / raw)
To: Jiri Slaby (SUSE)
Cc: gregkh, linux-serial, linux-kernel, Richard Henderson,
Ivan Kokshaysky, Matt Turner, linux-alpha
[-- Attachment #1: Type: text/plain, Size: 1679 bytes --]
On Mon, 27 Nov 2023, Jiri Slaby (SUSE) wrote:
> 'need_cr' is a flag, so type it properly to be a 'bool'. Move the
> declaration into the loop too. That ensures the variable is initialized
> properly even if the code was moved somehow.
>
> Signed-off-by: Jiri Slaby (SUSE) <jirislaby@kernel.org>
> Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
> Cc: Ivan Kokshaysky <ink@jurassic.park.msu.ru>
> Cc: Matt Turner <mattst88@gmail.com>
> Cc: linux-alpha@vger.kernel.org
> ---
> arch/alpha/kernel/srmcons.c | 7 +++----
> 1 file changed, 3 insertions(+), 4 deletions(-)
>
> diff --git a/arch/alpha/kernel/srmcons.c b/arch/alpha/kernel/srmcons.c
> index 32bc098de7da..c6b821afbfd3 100644
> --- a/arch/alpha/kernel/srmcons.c
> +++ b/arch/alpha/kernel/srmcons.c
> @@ -94,17 +94,16 @@ srmcons_do_write(struct tty_port *port, const char *buf, int count)
> static char str_cr[1] = "\r";
> size_t c;
> srmcons_result result;
> - int need_cr;
>
> while (count > 0) {
> - need_cr = 0;
> + bool need_cr = false;
> /*
> * Break it up into reasonable size chunks to allow a chance
> * for input to get in
> */
> for (c = 0; c < min_t(size_t, 128U, count) && !need_cr; c++)
> if (buf[c] == '\n')
> - need_cr = 1;
> + need_cr = true;
>
> while (c > 0) {
> result.as_long = callback_puts(0, buf, c);
> @@ -122,7 +121,7 @@ srmcons_do_write(struct tty_port *port, const char *buf, int count)
> while (need_cr) {
> result.as_long = callback_puts(0, str_cr, 1);
> if (result.bits.c > 0)
> - need_cr = 0;
> + need_cr = false;
> }
> }
> }
>
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
--
i.
^ permalink raw reply [flat|nested] 8+ messages in thread
* [PATCH 5/5] tty: srmcons: remove 'str_cr' and use string directly
[not found] <20231127123713.14504-1-jirislaby@kernel.org>
` (2 preceding siblings ...)
2023-11-27 12:37 ` [PATCH 4/5] tty: srmcons: switch need_cr to bool Jiri Slaby (SUSE)
@ 2023-11-27 12:37 ` Jiri Slaby (SUSE)
2023-11-28 13:26 ` Ilpo Järvinen
3 siblings, 1 reply; 8+ messages in thread
From: Jiri Slaby (SUSE) @ 2023-11-27 12:37 UTC (permalink / raw)
To: gregkh
Cc: linux-serial, linux-kernel, Jiri Slaby (SUSE), Richard Henderson,
Ivan Kokshaysky, Matt Turner, linux-alpha
'str_cr' contains a single character: \r. There is no need to declare it
as array. Instead, pass the character (as a string) to callback_puts()
directly. This ensures the string is in proper .rodata (const) section
and makes the code more obvious.
Signed-off-by: Jiri Slaby (SUSE) <jirislaby@kernel.org>
Cc: Richard Henderson <richard.henderson@linaro.org>
Cc: Ivan Kokshaysky <ink@jurassic.park.msu.ru>
Cc: Matt Turner <mattst88@gmail.com>
Cc: linux-alpha@vger.kernel.org
---
Notes:
[v2]
pass "\r" directly to callback_puts() as Richard suggests
spell correct \r in the commit log as Ilpo noticed
arch/alpha/kernel/srmcons.c | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
diff --git a/arch/alpha/kernel/srmcons.c b/arch/alpha/kernel/srmcons.c
index c6b821afbfd3..42deea53beab 100644
--- a/arch/alpha/kernel/srmcons.c
+++ b/arch/alpha/kernel/srmcons.c
@@ -91,7 +91,6 @@ srmcons_receive_chars(struct timer_list *t)
static void
srmcons_do_write(struct tty_port *port, const char *buf, int count)
{
- static char str_cr[1] = "\r";
size_t c;
srmcons_result result;
@@ -119,7 +118,7 @@ srmcons_do_write(struct tty_port *port, const char *buf, int count)
}
while (need_cr) {
- result.as_long = callback_puts(0, str_cr, 1);
+ result.as_long = callback_puts(0, "\r", 1);
if (result.bits.c > 0)
need_cr = false;
}
--
2.42.1
^ permalink raw reply related [flat|nested] 8+ messages in thread* Re: [PATCH 5/5] tty: srmcons: remove 'str_cr' and use string directly
2023-11-27 12:37 ` [PATCH 5/5] tty: srmcons: remove 'str_cr' and use string directly Jiri Slaby (SUSE)
@ 2023-11-28 13:26 ` Ilpo Järvinen
0 siblings, 0 replies; 8+ messages in thread
From: Ilpo Järvinen @ 2023-11-28 13:26 UTC (permalink / raw)
To: Jiri Slaby (SUSE)
Cc: gregkh, linux-serial, linux-kernel, Richard Henderson,
Ivan Kokshaysky, Matt Turner, linux-alpha
[-- Attachment #1: Type: text/plain, Size: 1569 bytes --]
On Mon, 27 Nov 2023, Jiri Slaby (SUSE) wrote:
> 'str_cr' contains a single character: \r. There is no need to declare it
> as array. Instead, pass the character (as a string) to callback_puts()
> directly. This ensures the string is in proper .rodata (const) section
> and makes the code more obvious.
>
> Signed-off-by: Jiri Slaby (SUSE) <jirislaby@kernel.org>
> Cc: Richard Henderson <richard.henderson@linaro.org>
> Cc: Ivan Kokshaysky <ink@jurassic.park.msu.ru>
> Cc: Matt Turner <mattst88@gmail.com>
> Cc: linux-alpha@vger.kernel.org
> ---
>
> Notes:
> [v2]
> pass "\r" directly to callback_puts() as Richard suggests
> spell correct \r in the commit log as Ilpo noticed
>
> arch/alpha/kernel/srmcons.c | 3 +--
> 1 file changed, 1 insertion(+), 2 deletions(-)
>
> diff --git a/arch/alpha/kernel/srmcons.c b/arch/alpha/kernel/srmcons.c
> index c6b821afbfd3..42deea53beab 100644
> --- a/arch/alpha/kernel/srmcons.c
> +++ b/arch/alpha/kernel/srmcons.c
> @@ -91,7 +91,6 @@ srmcons_receive_chars(struct timer_list *t)
> static void
> srmcons_do_write(struct tty_port *port, const char *buf, int count)
> {
> - static char str_cr[1] = "\r";
> size_t c;
> srmcons_result result;
>
> @@ -119,7 +118,7 @@ srmcons_do_write(struct tty_port *port, const char *buf, int count)
> }
>
> while (need_cr) {
> - result.as_long = callback_puts(0, str_cr, 1);
> + result.as_long = callback_puts(0, "\r", 1);
> if (result.bits.c > 0)
> need_cr = false;
> }
>
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
--
i.
^ permalink raw reply [flat|nested] 8+ messages in thread