The function converts strings like ttyS0 and ttyUSB0 to dev_t like
(4, 64) and (188, 0). Subsequent patch in this set will call it to
convert user-supplied device into device number. The function does
some basic sanity checks on the string passed in. It currently supports
ttyS*, ttyUSB* and, for selected synths, lp*.
In order to do this, the patch also introduces a string member variable
named 'dev' to struct spk_synth. 'dev' represents the device name -
ttyUSB0 etc - which needs conversion to dev_t.
Signed-off-by: Okash Khawaja <[email protected]>
Reviewed-by: Samuel Thibault <[email protected]>
---
drivers/staging/speakup/spk_priv.h | 2
drivers/staging/speakup/spk_ttyio.c | 105 ++++++++++++++++++++++++++++++++++++
drivers/staging/speakup/spk_types.h | 1
3 files changed, 108 insertions(+)
--- a/drivers/staging/speakup/spk_priv.h
+++ b/drivers/staging/speakup/spk_priv.h
@@ -40,6 +40,8 @@
#define KT_SPKUP 15
#define SPK_SYNTH_TIMEOUT 100000 /* in micro-seconds */
+#define SYNTH_DEFAULT_DEV "ttyS0"
+#define SYNTH_DEFAULT_SER 0
const struct old_serial_port *spk_serial_init(int index);
void spk_stop_serial_interrupt(void);
--- a/drivers/staging/speakup/spk_ttyio.c
+++ b/drivers/staging/speakup/spk_ttyio.c
@@ -7,6 +7,13 @@
#include "spk_types.h"
#include "spk_priv.h"
+/* Supported device types */
+#define DEV_PREFIX_TTYS "ttyS"
+#define DEV_PREFIX_TTYUSB "ttyUSB"
+#define DEV_PREFIX_LP "lp"
+
+const char *lp_supported[] = { "acntsa", "bns", "dummy", "txprt" };
+
struct spk_ldisc_data {
char buf;
struct semaphore sem;
@@ -16,6 +23,104 @@ struct spk_ldisc_data {
static struct spk_synth *spk_ttyio_synth;
static struct tty_struct *speakup_tty;
+static int name_to_dev(const char *name, dev_t *dev_no)
+{
+ int maj = -1, min = -1;
+
+ if (strncmp(name, DEV_PREFIX_TTYS, strlen(DEV_PREFIX_TTYS)) == 0) {
+ if (kstrtoint(name + strlen(DEV_PREFIX_TTYS), 10, &min)) {
+ pr_err("speakup: Invalid ser param. Must be \
+ between 0 and 191 inclusive.\n");
+ return -EINVAL;
+ }
+ maj = 4;
+
+ if (min < 0 || min > 191) {
+ pr_err("speakup: Invalid ser param. Must be \
+ between 0 and 191 inclusive.\n");
+ return -EINVAL;
+ }
+ min = min + 64;
+ } else if (strncmp(name, DEV_PREFIX_TTYUSB, strlen(DEV_PREFIX_TTYUSB))
+ == 0) {
+ if (kstrtoint(name + strlen(DEV_PREFIX_TTYUSB), 10, &min)) {
+ pr_err("speakup: Invalid ttyUSB number. \
+ Must be a number from 0 onwards\n");
+ return -EINVAL;
+ }
+ maj = 188;
+
+ if (min < 0) {
+ pr_err("speakup: Invalid ttyUSB number. \
+ Must be a number from 0 onwards\n");
+ return -EINVAL;
+ }
+ } else if (strncmp(name, DEV_PREFIX_LP, strlen(DEV_PREFIX_LP)) == 0) {
+ if (kstrtoint(name + strlen(DEV_PREFIX_LP), 10, &min)) {
+ pr_warn("speakup: Invalid lp number. \
+ Must be a number from 0 onwards\n");
+ return -EINVAL;
+ }
+ maj = 6;
+
+ if (min < 0) {
+ pr_warn("speakup: Invalid lp number. \
+ Must be a number from 0 onwards\n");
+ return -EINVAL;
+ }
+ }
+
+ if (maj == -1 || min == -1)
+ return -EINVAL;
+
+ /* if here, maj and min must be valid */
+ *dev_no = MKDEV(maj, min);
+
+ return 0;
+}
+
+int ser_to_dev(int ser, dev_t *dev_no)
+{
+ if (ser < 0 || ser > (255 - 64)) {
+ pr_err("speakup: Invalid ser param. \
+ Must be between 0 and 191 inclusive.\n");
+
+ return -EINVAL;
+ }
+
+ *dev_no = MKDEV(4, (64 + ser));
+ return 0;
+}
+
+static int get_dev_to_use(struct spk_synth *synth, dev_t *dev_no)
+{
+ /* use ser only when dev is not specified */
+ if (strcmp(synth->dev, SYNTH_DEFAULT_DEV) || synth->ser == SYNTH_DEFAULT_SER) {
+ /* for /dev/lp* check if synth is supported */
+ if (strncmp(synth->dev, DEV_PREFIX_LP, strlen(DEV_PREFIX_LP)) == 0) {
+ int i;
+
+ for (i = 0; i < ARRAY_SIZE(lp_supported); i++) {
+ if (strcmp(synth->name, lp_supported[i]) == 0)
+ break;
+ }
+
+ if (i >= ARRAY_SIZE(lp_supported)) {
+ pr_err("speakup: lp* is only supported on:");
+ for (i = 0; i < ARRAY_SIZE(lp_supported); i++)
+ pr_cont(" %s", lp_supported[i]);
+ pr_cont("\n");
+
+ return -ENOTSUPP;
+ }
+ }
+
+ return name_to_dev(synth->dev, dev_no);
+ }
+
+ return ser_to_dev(synth->ser, dev_no);
+}
+
static int spk_ttyio_ldisc_open(struct tty_struct *tty)
{
struct spk_ldisc_data *ldisc_data;
--- a/drivers/staging/speakup/spk_types.h
+++ b/drivers/staging/speakup/spk_types.h
@@ -169,6 +169,7 @@ struct spk_synth {
int jiffies;
int full;
int ser;
+ char *dev;
short flags;
short startup;
const int checkval; /* for validating a proper synth module */
On Tue, Jun 13, 2017 at 11:37:03PM +0100, [email protected] wrote:
> The function converts strings like ttyS0 and ttyUSB0 to dev_t like
> (4, 64) and (188, 0). Subsequent patch in this set will call it to
> convert user-supplied device into device number. The function does
> some basic sanity checks on the string passed in. It currently supports
> ttyS*, ttyUSB* and, for selected synths, lp*.
>
> In order to do this, the patch also introduces a string member variable
> named 'dev' to struct spk_synth. 'dev' represents the device name -
> ttyUSB0 etc - which needs conversion to dev_t.
>
> Signed-off-by: Okash Khawaja <[email protected]>
> Reviewed-by: Samuel Thibault <[email protected]>
>
> ---
> drivers/staging/speakup/spk_priv.h | 2
> drivers/staging/speakup/spk_ttyio.c | 105 ++++++++++++++++++++++++++++++++++++
> drivers/staging/speakup/spk_types.h | 1
> 3 files changed, 108 insertions(+)
>
> --- a/drivers/staging/speakup/spk_priv.h
> +++ b/drivers/staging/speakup/spk_priv.h
> @@ -40,6 +40,8 @@
>
> #define KT_SPKUP 15
> #define SPK_SYNTH_TIMEOUT 100000 /* in micro-seconds */
> +#define SYNTH_DEFAULT_DEV "ttyS0"
> +#define SYNTH_DEFAULT_SER 0
>
> const struct old_serial_port *spk_serial_init(int index);
> void spk_stop_serial_interrupt(void);
> --- a/drivers/staging/speakup/spk_ttyio.c
> +++ b/drivers/staging/speakup/spk_ttyio.c
> @@ -7,6 +7,13 @@
> #include "spk_types.h"
> #include "spk_priv.h"
>
> +/* Supported device types */
> +#define DEV_PREFIX_TTYS "ttyS"
> +#define DEV_PREFIX_TTYUSB "ttyUSB"
> +#define DEV_PREFIX_LP "lp"
> +
> +const char *lp_supported[] = { "acntsa", "bns", "dummy", "txprt" };
> +
> struct spk_ldisc_data {
> char buf;
> struct semaphore sem;
> @@ -16,6 +23,104 @@ struct spk_ldisc_data {
> static struct spk_synth *spk_ttyio_synth;
> static struct tty_struct *speakup_tty;
>
> +static int name_to_dev(const char *name, dev_t *dev_no)
> +{
> + int maj = -1, min = -1;
> +
> + if (strncmp(name, DEV_PREFIX_TTYS, strlen(DEV_PREFIX_TTYS)) == 0) {
> + if (kstrtoint(name + strlen(DEV_PREFIX_TTYS), 10, &min)) {
> + pr_err("speakup: Invalid ser param. Must be \
> + between 0 and 191 inclusive.\n");
String needs fixed.
> + return -EINVAL;
Preserve the error code from kstrtoint().
> + }
> + maj = 4;
> +
> + if (min < 0 || min > 191) {
> + pr_err("speakup: Invalid ser param. Must be \
> + between 0 and 191 inclusive.\n");
This too.
> + return -EINVAL;
> + }
> + min = min + 64;
> + } else if (strncmp(name, DEV_PREFIX_TTYUSB, strlen(DEV_PREFIX_TTYUSB))
> + == 0) {
> + if (kstrtoint(name + strlen(DEV_PREFIX_TTYUSB), 10, &min)) {
> + pr_err("speakup: Invalid ttyUSB number. \
> + Must be a number from 0 onwards\n");
> + return -EINVAL;
Same.
> + }
> + maj = 188;
> +
> + if (min < 0) {
> + pr_err("speakup: Invalid ttyUSB number. \
> + Must be a number from 0 onwards\n");
Same.
> + return -EINVAL;
> + }
> + } else if (strncmp(name, DEV_PREFIX_LP, strlen(DEV_PREFIX_LP)) == 0) {
> + if (kstrtoint(name + strlen(DEV_PREFIX_LP), 10, &min)) {
> + pr_warn("speakup: Invalid lp number. \
> + Must be a number from 0 onwards\n");
> + return -EINVAL;
Same. Preserve.
> + }
> + maj = 6;
> +
> + if (min < 0) {
> + pr_warn("speakup: Invalid lp number. \
> + Must be a number from 0 onwards\n");
Again.
> + return -EINVAL;
> + }
> + }
> +
> + if (maj == -1 || min == -1)
> + return -EINVAL;
> +
> + /* if here, maj and min must be valid */
> + *dev_no = MKDEV(maj, min);
> +
> + return 0;
> +}
> +
> +int ser_to_dev(int ser, dev_t *dev_no)
> +{
> + if (ser < 0 || ser > (255 - 64)) {
> + pr_err("speakup: Invalid ser param. \
> + Must be between 0 and 191 inclusive.\n");
String. I feel like 191 and 255 - 64 are the same. Let's just use 191
everywhere.
> +
> + return -EINVAL;
> + }
> +
> + *dev_no = MKDEV(4, (64 + ser));
> + return 0;
> +}
> +
> +static int get_dev_to_use(struct spk_synth *synth, dev_t *dev_no)
> +{
> + /* use ser only when dev is not specified */
> + if (strcmp(synth->dev, SYNTH_DEFAULT_DEV) || synth->ser == SYNTH_DEFAULT_SER) {
> + /* for /dev/lp* check if synth is supported */
> + if (strncmp(synth->dev, DEV_PREFIX_LP, strlen(DEV_PREFIX_LP)) == 0) {
> + int i;
> +
> + for (i = 0; i < ARRAY_SIZE(lp_supported); i++) {
> + if (strcmp(synth->name, lp_supported[i]) == 0)
> + break;
> + }
> +
> + if (i >= ARRAY_SIZE(lp_supported)) {
> + pr_err("speakup: lp* is only supported on:");
> + for (i = 0; i < ARRAY_SIZE(lp_supported); i++)
> + pr_cont(" %s", lp_supported[i]);
> + pr_cont("\n");
> +
> + return -ENOTSUPP;
> + }
> + }
> +
> + return name_to_dev(synth->dev, dev_no);
> + }
> +
> + return ser_to_dev(synth->ser, dev_no);
> +}
> +
> static int spk_ttyio_ldisc_open(struct tty_struct *tty)
> {
> struct spk_ldisc_data *ldisc_data;
> --- a/drivers/staging/speakup/spk_types.h
> +++ b/drivers/staging/speakup/spk_types.h
> @@ -169,6 +169,7 @@ struct spk_synth {
> int jiffies;
> int full;
> int ser;
> + char *dev;
Could you call it "dev_name" instead? I normally expect "dev" to be a
device struct.
regards,
dan carpenter
On Tue, Jun 13, 2017 at 11:37:03PM +0100, [email protected] wrote:
> The function converts strings like ttyS0 and ttyUSB0 to dev_t like
> (4, 64) and (188, 0). Subsequent patch in this set will call it to
> convert user-supplied device into device number. The function does
> some basic sanity checks on the string passed in. It currently supports
> ttyS*, ttyUSB* and, for selected synths, lp*.
>
> In order to do this, the patch also introduces a string member variable
> named 'dev' to struct spk_synth. 'dev' represents the device name -
> ttyUSB0 etc - which needs conversion to dev_t.
>
> Signed-off-by: Okash Khawaja <[email protected]>
> Reviewed-by: Samuel Thibault <[email protected]>
>
> ---
> drivers/staging/speakup/spk_priv.h | 2
> drivers/staging/speakup/spk_ttyio.c | 105 ++++++++++++++++++++++++++++++++++++
> drivers/staging/speakup/spk_types.h | 1
> 3 files changed, 108 insertions(+)
>
> --- a/drivers/staging/speakup/spk_priv.h
> +++ b/drivers/staging/speakup/spk_priv.h
> @@ -40,6 +40,8 @@
>
> #define KT_SPKUP 15
> #define SPK_SYNTH_TIMEOUT 100000 /* in micro-seconds */
> +#define SYNTH_DEFAULT_DEV "ttyS0"
> +#define SYNTH_DEFAULT_SER 0
>
> const struct old_serial_port *spk_serial_init(int index);
> void spk_stop_serial_interrupt(void);
> --- a/drivers/staging/speakup/spk_ttyio.c
> +++ b/drivers/staging/speakup/spk_ttyio.c
> @@ -7,6 +7,13 @@
> #include "spk_types.h"
> #include "spk_priv.h"
>
> +/* Supported device types */
> +#define DEV_PREFIX_TTYS "ttyS"
> +#define DEV_PREFIX_TTYUSB "ttyUSB"
> +#define DEV_PREFIX_LP "lp"
> +
> +const char *lp_supported[] = { "acntsa", "bns", "dummy", "txprt" };
> +
> struct spk_ldisc_data {
> char buf;
> struct semaphore sem;
> @@ -16,6 +23,104 @@ struct spk_ldisc_data {
> static struct spk_synth *spk_ttyio_synth;
> static struct tty_struct *speakup_tty;
>
> +static int name_to_dev(const char *name, dev_t *dev_no)
> +{
> + int maj = -1, min = -1;
> +
> + if (strncmp(name, DEV_PREFIX_TTYS, strlen(DEV_PREFIX_TTYS)) == 0) {
> + if (kstrtoint(name + strlen(DEV_PREFIX_TTYS), 10, &min)) {
> + pr_err("speakup: Invalid ser param. Must be \
> + between 0 and 191 inclusive.\n");
> + return -EINVAL;
> + }
> + maj = 4;
> +
> + if (min < 0 || min > 191) {
> + pr_err("speakup: Invalid ser param. Must be \
> + between 0 and 191 inclusive.\n");
> + return -EINVAL;
> + }
> + min = min + 64;
> + } else if (strncmp(name, DEV_PREFIX_TTYUSB, strlen(DEV_PREFIX_TTYUSB))
> + == 0) {
> + if (kstrtoint(name + strlen(DEV_PREFIX_TTYUSB), 10, &min)) {
> + pr_err("speakup: Invalid ttyUSB number. \
> + Must be a number from 0 onwards\n");
> + return -EINVAL;
> + }
> + maj = 188;
> +
> + if (min < 0) {
> + pr_err("speakup: Invalid ttyUSB number. \
> + Must be a number from 0 onwards\n");
> + return -EINVAL;
> + }
> + } else if (strncmp(name, DEV_PREFIX_LP, strlen(DEV_PREFIX_LP)) == 0) {
> + if (kstrtoint(name + strlen(DEV_PREFIX_LP), 10, &min)) {
> + pr_warn("speakup: Invalid lp number. \
> + Must be a number from 0 onwards\n");
> + return -EINVAL;
> + }
> + maj = 6;
> +
> + if (min < 0) {
> + pr_warn("speakup: Invalid lp number. \
> + Must be a number from 0 onwards\n");
> + return -EINVAL;
> + }
> + }
> +
> + if (maj == -1 || min == -1)
> + return -EINVAL;
> +
> + /* if here, maj and min must be valid */
> + *dev_no = MKDEV(maj, min);
> +
> + return 0;
> +}
Eeek, no, let's never try to parse strings like this and "figure out"
what major/minor number it is. That's madness and will break if we ever
make all char majors dynamic (there's a thread on lkml about that.)
Why would the kernel need to know major/minor? Is it going to open a
device node? If so, again, crazy stuff, that's not good...
thanks,
greg k-h
On Wed, Jun 14, 2017 at 12:13:26PM +0200, Greg Kroah-Hartman wrote:
> On Tue, Jun 13, 2017 at 11:37:03PM +0100, [email protected] wrote:
> > The function converts strings like ttyS0 and ttyUSB0 to dev_t like
> > (4, 64) and (188, 0). Subsequent patch in this set will call it to
> > convert user-supplied device into device number. The function does
> > some basic sanity checks on the string passed in. It currently supports
> > ttyS*, ttyUSB* and, for selected synths, lp*.
> >
> > In order to do this, the patch also introduces a string member variable
> > named 'dev' to struct spk_synth. 'dev' represents the device name -
> > ttyUSB0 etc - which needs conversion to dev_t.
> >
> > Signed-off-by: Okash Khawaja <[email protected]>
> > Reviewed-by: Samuel Thibault <[email protected]>
> >
> > ---
> > drivers/staging/speakup/spk_priv.h | 2
> > drivers/staging/speakup/spk_ttyio.c | 105 ++++++++++++++++++++++++++++++++++++
> > drivers/staging/speakup/spk_types.h | 1
> > 3 files changed, 108 insertions(+)
> >
> > --- a/drivers/staging/speakup/spk_priv.h
> > +++ b/drivers/staging/speakup/spk_priv.h
> > @@ -40,6 +40,8 @@
> >
> > #define KT_SPKUP 15
> > #define SPK_SYNTH_TIMEOUT 100000 /* in micro-seconds */
> > +#define SYNTH_DEFAULT_DEV "ttyS0"
> > +#define SYNTH_DEFAULT_SER 0
> >
> > const struct old_serial_port *spk_serial_init(int index);
> > void spk_stop_serial_interrupt(void);
> > --- a/drivers/staging/speakup/spk_ttyio.c
> > +++ b/drivers/staging/speakup/spk_ttyio.c
> > @@ -7,6 +7,13 @@
> > #include "spk_types.h"
> > #include "spk_priv.h"
> >
> > +/* Supported device types */
> > +#define DEV_PREFIX_TTYS "ttyS"
> > +#define DEV_PREFIX_TTYUSB "ttyUSB"
> > +#define DEV_PREFIX_LP "lp"
> > +
> > +const char *lp_supported[] = { "acntsa", "bns", "dummy", "txprt" };
> > +
> > struct spk_ldisc_data {
> > char buf;
> > struct semaphore sem;
> > @@ -16,6 +23,104 @@ struct spk_ldisc_data {
> > static struct spk_synth *spk_ttyio_synth;
> > static struct tty_struct *speakup_tty;
> >
> > +static int name_to_dev(const char *name, dev_t *dev_no)
> > +{
> > + int maj = -1, min = -1;
> > +
> > + if (strncmp(name, DEV_PREFIX_TTYS, strlen(DEV_PREFIX_TTYS)) == 0) {
> > + if (kstrtoint(name + strlen(DEV_PREFIX_TTYS), 10, &min)) {
> > + pr_err("speakup: Invalid ser param. Must be \
> > + between 0 and 191 inclusive.\n");
> > + return -EINVAL;
> > + }
> > + maj = 4;
> > +
> > + if (min < 0 || min > 191) {
> > + pr_err("speakup: Invalid ser param. Must be \
> > + between 0 and 191 inclusive.\n");
> > + return -EINVAL;
> > + }
> > + min = min + 64;
> > + } else if (strncmp(name, DEV_PREFIX_TTYUSB, strlen(DEV_PREFIX_TTYUSB))
> > + == 0) {
> > + if (kstrtoint(name + strlen(DEV_PREFIX_TTYUSB), 10, &min)) {
> > + pr_err("speakup: Invalid ttyUSB number. \
> > + Must be a number from 0 onwards\n");
> > + return -EINVAL;
> > + }
> > + maj = 188;
> > +
> > + if (min < 0) {
> > + pr_err("speakup: Invalid ttyUSB number. \
> > + Must be a number from 0 onwards\n");
> > + return -EINVAL;
> > + }
> > + } else if (strncmp(name, DEV_PREFIX_LP, strlen(DEV_PREFIX_LP)) == 0) {
> > + if (kstrtoint(name + strlen(DEV_PREFIX_LP), 10, &min)) {
> > + pr_warn("speakup: Invalid lp number. \
> > + Must be a number from 0 onwards\n");
> > + return -EINVAL;
> > + }
> > + maj = 6;
> > +
> > + if (min < 0) {
> > + pr_warn("speakup: Invalid lp number. \
> > + Must be a number from 0 onwards\n");
> > + return -EINVAL;
> > + }
> > + }
> > +
> > + if (maj == -1 || min == -1)
> > + return -EINVAL;
> > +
> > + /* if here, maj and min must be valid */
> > + *dev_no = MKDEV(maj, min);
> > +
> > + return 0;
> > +}
>
> Eeek, no, let's never try to parse strings like this and "figure out"
> what major/minor number it is. That's madness and will break if we ever
> make all char majors dynamic (there's a thread on lkml about that.)
>
> Why would the kernel need to know major/minor? Is it going to open a
> device node? If so, again, crazy stuff, that's not good...
Ah, no, nevermind, you just need the major/minor. So why not just take
that as the input here, and not a string?
thanks,
greg k-h
Greg KH, on mer. 14 juin 2017 12:15:41 +0200, wrote:
> Ah, no, nevermind, you just need the major/minor. So why not just take
> that as the input here, and not a string?
Because real users don't know about major/minor numbers.
Samuel
On Wed, Jun 14, 2017 at 01:26:09PM +0200, Samuel Thibault wrote:
> Greg KH, on mer. 14 juin 2017 12:15:41 +0200, wrote:
> > Ah, no, nevermind, you just need the major/minor. So why not just take
> > that as the input here, and not a string?
>
> Because real users don't know about major/minor numbers.
I'm not disagreeing, it's just that the kernel doesn't know about
"device names" either :)
And trying to have the kernel do the mapping based on strings like this
is not ok, sorry.
greg k-h
Greg KH, on mer. 14 juin 2017 13:48:02 +0200, wrote:
> On Wed, Jun 14, 2017 at 01:26:09PM +0200, Samuel Thibault wrote:
> > Greg KH, on mer. 14 juin 2017 12:15:41 +0200, wrote:
> > > Ah, no, nevermind, you just need the major/minor. So why not just take
> > > that as the input here, and not a string?
> >
> > Because real users don't know about major/minor numbers.
>
> I'm not disagreeing, it's just that the kernel doesn't know about
> "device names" either :)
Well, it does for the console= parameter, for instance. I know it's not
actually related with major/minor numbering, but it's still meant to
designate the same thing.
> And trying to have the kernel do the mapping based on strings like this
> is not ok, sorry.
So what is the solution? Users would find it completely crazy to have
to provide major/minor numbers.
Samuel
On Wed, Jun 14, 2017 at 3:18 PM, Samuel Thibault
<[email protected]> wrote:
> Greg KH, on mer. 14 juin 2017 13:48:02 +0200, wrote:
>> On Wed, Jun 14, 2017 at 01:26:09PM +0200, Samuel Thibault wrote:
>> > Greg KH, on mer. 14 juin 2017 12:15:41 +0200, wrote:
>> And trying to have the kernel do the mapping based on strings like this
>> is not ok, sorry.
>
> So what is the solution? Users would find it completely crazy to have
> to provide major/minor numbers.
Shouldn't udev take care of major/minor and alike stuff in user space?
--
With Best Regards,
Andy Shevchenko
Andy Shevchenko, on mer. 14 juin 2017 15:36:30 +0300, wrote:
> On Wed, Jun 14, 2017 at 3:18 PM, Samuel Thibault
> <[email protected]> wrote:
> > Greg KH, on mer. 14 juin 2017 13:48:02 +0200, wrote:
> >> On Wed, Jun 14, 2017 at 01:26:09PM +0200, Samuel Thibault wrote:
> >> > Greg KH, on mer. 14 juin 2017 12:15:41 +0200, wrote:
>
> >> And trying to have the kernel do the mapping based on strings like this
> >> is not ok, sorry.
> >
> > So what is the solution? Users would find it completely crazy to have
> > to provide major/minor numbers.
>
> Shouldn't udev take care of major/minor and alike stuff in user space?
We want speakup to get initialized before udev etc., just like the early
console.
Samuel
On Wed, Jun 14, 2017 at 03:36:30PM +0300, Andy Shevchenko wrote:
> On Wed, Jun 14, 2017 at 3:18 PM, Samuel Thibault
> <[email protected]> wrote:
> > Greg KH, on mer. 14 juin 2017 13:48:02 +0200, wrote:
> >> On Wed, Jun 14, 2017 at 01:26:09PM +0200, Samuel Thibault wrote:
> >> > Greg KH, on mer. 14 juin 2017 12:15:41 +0200, wrote:
>
> >> And trying to have the kernel do the mapping based on strings like this
> >> is not ok, sorry.
> >
> > So what is the solution? Users would find it completely crazy to have
> > to provide major/minor numbers.
>
> Shouldn't udev take care of major/minor and alike stuff in user space?
No, that's all handled by the kernel now, in devtmpfs.
The console stuff is odd though, but that is each driver defining the
name for itself, major/minor does not apply. Also, a separate driver is
not having to figure this all out. I wonder if we could just add a tty
core function for this as it does know the name of everything that has
been registered with it, right?
thanks,
greg k-h
Hi,
On Wed, Jun 14, 2017 at 03:04:18PM +0200, Greg Kroah-Hartman wrote:
> The console stuff is odd though, but that is each driver defining the
> name for itself, major/minor does not apply. Also, a separate driver is
> not having to figure this all out. I wonder if we could just add a tty
> core function for this as it does know the name of everything that has
> been registered with it, right?
I can start working on a patch for this. I'm still new to the tty code
so will follow up with any questions I might have.
Thanks,
Okash
Hi,
On Wed, Jun 14, 2017 at 9:23 AM, Dan Carpenter <[email protected]> wrote:
[...]
>
> Could you call it "dev_name" instead? I normally expect "dev" to be a
> device struct.
Thanks for the feedback. Will keep these in mind for next version of the patch.
Okash
Hi,
On Thu, Jun 15, 2017 at 07:52:51AM +0100, Okash Khawaja wrote:
> On Wed, Jun 14, 2017 at 03:04:18PM +0200, Greg Kroah-Hartman wrote:
> > The console stuff is odd though, but that is each driver defining the
> > name for itself, major/minor does not apply. Also, a separate driver is
> > not having to figure this all out. I wonder if we could just add a tty
> > core function for this as it does know the name of everything that has
> > been registered with it, right?
>
> I can start working on a patch for this. I'm still new to the tty code
> so will follow up with any questions I might have.
I've put together this function for converting device name to number.
It traverses tty_drivers list looking for corresponding dev name and
index. Is this the right approach?
Thanks,
Okash
On Sat, Jun 17, 2017 at 11:16:44AM +0100, Okash Khawaja wrote:
> Hi,
>
> On Thu, Jun 15, 2017 at 07:52:51AM +0100, Okash Khawaja wrote:
> > On Wed, Jun 14, 2017 at 03:04:18PM +0200, Greg Kroah-Hartman wrote:
> > > The console stuff is odd though, but that is each driver defining the
> > > name for itself, major/minor does not apply. Also, a separate driver is
> > > not having to figure this all out. I wonder if we could just add a tty
> > > core function for this as it does know the name of everything that has
> > > been registered with it, right?
> >
> > I can start working on a patch for this. I'm still new to the tty code
> > so will follow up with any questions I might have.
>
> I've put together this function for converting device name to number.
> It traverses tty_drivers list looking for corresponding dev name and
> index. Is this the right approach?
Yes, this looks great, nice job! It is a lot simpler than your previous
function, and now you are not limited to just a subset off the different
serial ports in the system.
Keep up the good work, it's really appreciated.
greg k-h
On Sat, Jun 17, 2017 at 12:25:34PM +0200, Greg Kroah-Hartman wrote:
> On Sat, Jun 17, 2017 at 11:16:44AM +0100, Okash Khawaja wrote:
> > Hi,
> >
> > On Thu, Jun 15, 2017 at 07:52:51AM +0100, Okash Khawaja wrote:
> > > On Wed, Jun 14, 2017 at 03:04:18PM +0200, Greg Kroah-Hartman wrote:
> > > > The console stuff is odd though, but that is each driver defining the
> > > > name for itself, major/minor does not apply. Also, a separate driver is
> > > > not having to figure this all out. I wonder if we could just add a tty
> > > > core function for this as it does know the name of everything that has
> > > > been registered with it, right?
> > >
> > > I can start working on a patch for this. I'm still new to the tty code
> > > so will follow up with any questions I might have.
> >
> > I've put together this function for converting device name to number.
> > It traverses tty_drivers list looking for corresponding dev name and
> > index. Is this the right approach?
>
> Yes, this looks great, nice job! It is a lot simpler than your previous
> function, and now you are not limited to just a subset off the different
> serial ports in the system.
>
> Keep up the good work, it's really appreciated.
Great, thanks. I'll re-send the full patch set as v2.
Okash