Re: [RFC PATCH] use snprintf instead of sprintf in rcu_torture_printk
On Thu, 2014-06-19 at 21:54 -0700, Paul E. McKenney wrote:
> On Thu, Jun 19, 2014 at 08:13:22PM -0400, Pranith Kumar wrote:
> > On 06/19/2014 07:49 PM, Paul E. McKenney wrote:
> > > On Thu, Jun 19, 2014 at 07:24:48PM -0400, Pranith Kumar wrote:
> > >> (dropping some CCs)
> > >>
> > >> On 06/19/2014 05:00 PM, Paul E. McKenney wrote:
> > >>> On Wed, Jun 18, 2014 at 12:49:42PM -0700, Joe Perches wrote:
> >
> > I believe the function doesn't work well.
> >
> > static void
> > rcu_torture_stats_print(void)
> > {
> > int size = nr_cpu_ids * 200 + 8192;
> > char *buf;
> >
> > buf = kmalloc(size, GFP_KERNEL);
> > if (!buf) {
> > pr_err("rcu-torture: Out of memory, need: %d\n", size);
> > return;
> > }
> > rcu_torture_printk(buf);
> > pr_alert("%s", buf);
> > kfree(buf);
> > }
> >
> > rcu_torture_printk simply fills buf
> >
> > btw: I believe the arguments should pass size and
> > rcu_torture_printk should use snprintf/size
> >
> > but all printks are limited to a maximum of 1024
> > bytes so the large allocation is senseless and
> > would even if it worked, would likely need to be
> > vmalloc/vfree
> > >>>
> > >>> Fair point!
> > >>>
> > >>> Pranith, Romanov, if you would like part of RCU that is less touchy
> > >>> about random hacking, this would be a good place to start. Scripts in
> > >>> tools/testing/selftests/rcutorture/bin do care about some of the format,
> > >>> but the variable-length portion generated by cur_ops->stats() is as far
> > >>> as I know only parsed by human eyes.
> > >>>
> > >>
> > >> Here is a first run of the change. Please let me know if I am totally
> > >> off. RFC. :)
> > >
> > > Thank you for taking this on!
> >
> > You are most welcome :)
> >
> > >
> > >> Three things on Todo list:
> > >>
> > >> * We need to check that we are using less than the allocated size of the
> > >> buffer (used > size). (we are allocating a big buffer, so not sure if
> > >> necessary)
> > >> * Need to check with the scripts if they are working.
> > >> * I used a loop for pr_alert(). I am not sure this is right, there
> > >> should be a better way for printing large buffers
> > >>
> > >> If the overall structure is ok I will go ahead and check how the scripts
> > >> are handling these changes.
> > >
> > > One other thing... Convince this function (and a few others that it
> > > calls) that the system really has 4096 CPUs, run this code, and see wshat
> > > actually happens both before and after. Just to get a bit of practice
> > > mixed in with the theory. ;-)
> > >
> >
> > OK. I think to do this I need to use 4096 instead of nr_cpu_ids. I will try
> > this and see how it goes :)
>
> That's the spirit! You will probably have to adjust a few other things
> as well.
Perhaps this should not use the allocation / page buffer
and instead call pr_alert and pr_cont directly like this:
include/linux/torture.h | 2 +-
kernel/rcu/rcutorture.c | 110 +++-
kernel/torture.c| 16 +++
3 files changed, 60 insertions(+), 68 deletions(-)
diff --git a/include/linux/torture.h b/include/linux/torture.h
index 5ca58fc..fec46f8 100644
--- a/include/linux/torture.h
+++ b/include/linux/torture.h
@@ -51,7 +51,7 @@
/* Definitions for online/offline exerciser. */
int torture_onoff_init(long ooholdoff, long oointerval);
-char *torture_onoff_stats(char *page);
+void torture_onoff_stats(void);
bool torture_onoff_failures(void);
/* Low-rider random number generator. */
diff --git a/kernel/rcu/rcutorture.c b/kernel/rcu/rcutorture.c
index 7fa34f8..bb6db08 100644
--- a/kernel/rcu/rcutorture.c
+++ b/kernel/rcu/rcutorture.c
@@ -242,7 +242,7 @@ struct rcu_torture_ops {
void (*call)(struct rcu_head *head, void (*func)(struct rcu_head *rcu));
void (*cb_barrier)(void);
void (*fqs)(void);
- void (*stats)(char *page);
+ void (*stats)(void);
int irq_capable;
int can_boost;
const char *name;
@@ -525,21 +525,21 @@ static void srcu_torture_barrier(void)
srcu_barrier(_ctl);
}
-static void srcu_torture_stats(char *page)
+static void srcu_torture_stats(void)
{
int cpu;
int idx = srcu_ctl.completed & 0x1;
- page += sprintf(page, "%s%s per-CPU(idx=%d):",
- torture_type, TORTURE_FLAG, idx);
+ pr_alert("%s%s per-CPU(idx=%d):",
+torture_type, TORTURE_FLAG, idx);
for_each_possible_cpu(cpu) {
long c0, c1;
c0 = (long)per_cpu_ptr(srcu_ctl.per_cpu_ref, cpu)->c[!idx];
c1 = (long)per_cpu_ptr(srcu_ctl.per_cpu_ref, cpu)->c[idx];
- page += sprintf(page, " %d(%ld,%ld)", cpu, c0, c1);
+ pr_cont(" %d(%ld,%ld)", cpu, c0, c1);
}
-
Re: [RFC PATCH] use snprintf instead of sprintf in rcu_torture_printk
On Thu, 2014-06-19 at 21:54 -0700, Paul E. McKenney wrote:
On Thu, Jun 19, 2014 at 08:13:22PM -0400, Pranith Kumar wrote:
On 06/19/2014 07:49 PM, Paul E. McKenney wrote:
On Thu, Jun 19, 2014 at 07:24:48PM -0400, Pranith Kumar wrote:
(dropping some CCs)
On 06/19/2014 05:00 PM, Paul E. McKenney wrote:
On Wed, Jun 18, 2014 at 12:49:42PM -0700, Joe Perches wrote:
I believe the function doesn't work well.
static void
rcu_torture_stats_print(void)
{
int size = nr_cpu_ids * 200 + 8192;
char *buf;
buf = kmalloc(size, GFP_KERNEL);
if (!buf) {
pr_err(rcu-torture: Out of memory, need: %d\n, size);
return;
}
rcu_torture_printk(buf);
pr_alert(%s, buf);
kfree(buf);
}
rcu_torture_printk simply fills buf
btw: I believe the arguments should pass size and
rcu_torture_printk should use snprintf/size
but all printks are limited to a maximum of 1024
bytes so the large allocation is senseless and
would even if it worked, would likely need to be
vmalloc/vfree
Fair point!
Pranith, Romanov, if you would like part of RCU that is less touchy
about random hacking, this would be a good place to start. Scripts in
tools/testing/selftests/rcutorture/bin do care about some of the format,
but the variable-length portion generated by cur_ops-stats() is as far
as I know only parsed by human eyes.
Here is a first run of the change. Please let me know if I am totally
off. RFC. :)
Thank you for taking this on!
You are most welcome :)
Three things on Todo list:
* We need to check that we are using less than the allocated size of the
buffer (used size). (we are allocating a big buffer, so not sure if
necessary)
* Need to check with the scripts if they are working.
* I used a loop for pr_alert(). I am not sure this is right, there
should be a better way for printing large buffers
If the overall structure is ok I will go ahead and check how the scripts
are handling these changes.
One other thing... Convince this function (and a few others that it
calls) that the system really has 4096 CPUs, run this code, and see wshat
actually happens both before and after. Just to get a bit of practice
mixed in with the theory. ;-)
OK. I think to do this I need to use 4096 instead of nr_cpu_ids. I will try
this and see how it goes :)
That's the spirit! You will probably have to adjust a few other things
as well.
Perhaps this should not use the allocation / page buffer
and instead call pr_alert and pr_cont directly like this:
include/linux/torture.h | 2 +-
kernel/rcu/rcutorture.c | 110 +++-
kernel/torture.c| 16 +++
3 files changed, 60 insertions(+), 68 deletions(-)
diff --git a/include/linux/torture.h b/include/linux/torture.h
index 5ca58fc..fec46f8 100644
--- a/include/linux/torture.h
+++ b/include/linux/torture.h
@@ -51,7 +51,7 @@
/* Definitions for online/offline exerciser. */
int torture_onoff_init(long ooholdoff, long oointerval);
-char *torture_onoff_stats(char *page);
+void torture_onoff_stats(void);
bool torture_onoff_failures(void);
/* Low-rider random number generator. */
diff --git a/kernel/rcu/rcutorture.c b/kernel/rcu/rcutorture.c
index 7fa34f8..bb6db08 100644
--- a/kernel/rcu/rcutorture.c
+++ b/kernel/rcu/rcutorture.c
@@ -242,7 +242,7 @@ struct rcu_torture_ops {
void (*call)(struct rcu_head *head, void (*func)(struct rcu_head *rcu));
void (*cb_barrier)(void);
void (*fqs)(void);
- void (*stats)(char *page);
+ void (*stats)(void);
int irq_capable;
int can_boost;
const char *name;
@@ -525,21 +525,21 @@ static void srcu_torture_barrier(void)
srcu_barrier(srcu_ctl);
}
-static void srcu_torture_stats(char *page)
+static void srcu_torture_stats(void)
{
int cpu;
int idx = srcu_ctl.completed 0x1;
- page += sprintf(page, %s%s per-CPU(idx=%d):,
- torture_type, TORTURE_FLAG, idx);
+ pr_alert(%s%s per-CPU(idx=%d):,
+torture_type, TORTURE_FLAG, idx);
for_each_possible_cpu(cpu) {
long c0, c1;
c0 = (long)per_cpu_ptr(srcu_ctl.per_cpu_ref, cpu)-c[!idx];
c1 = (long)per_cpu_ptr(srcu_ctl.per_cpu_ref, cpu)-c[idx];
- page += sprintf(page, %d(%ld,%ld), cpu, c0, c1);
+ pr_cont( %d(%ld,%ld), cpu, c0, c1);
}
- sprintf(page, \n);
+ pr_cont(\n);
}
static void srcu_torture_synchronize_expedited(void)
@@ -1031,10 +1031,10 @@ rcu_torture_reader(void *arg)
}
/*
- * Create an RCU-torture statistics message in the specified buffer.
+ * Print RCU-torture statistics
*/
static void
-rcu_torture_printk(char *page)
+rcu_torture_printk(void)
{
Re: [RFC PATCH] use snprintf instead of sprintf in rcu_torture_printk
On Thu, Jun 19, 2014 at 08:13:22PM -0400, Pranith Kumar wrote:
> On 06/19/2014 07:49 PM, Paul E. McKenney wrote:
> > On Thu, Jun 19, 2014 at 07:24:48PM -0400, Pranith Kumar wrote:
> >> (dropping some CCs)
> >>
> >> On 06/19/2014 05:00 PM, Paul E. McKenney wrote:
> >>> On Wed, Jun 18, 2014 at 12:49:42PM -0700, Joe Perches wrote:
>
> I believe the function doesn't work well.
>
> static void
> rcu_torture_stats_print(void)
> {
> int size = nr_cpu_ids * 200 + 8192;
> char *buf;
>
> buf = kmalloc(size, GFP_KERNEL);
> if (!buf) {
> pr_err("rcu-torture: Out of memory, need: %d\n", size);
> return;
> }
> rcu_torture_printk(buf);
> pr_alert("%s", buf);
> kfree(buf);
> }
>
> rcu_torture_printk simply fills buf
>
> btw: I believe the arguments should pass size and
> rcu_torture_printk should use snprintf/size
>
> but all printks are limited to a maximum of 1024
> bytes so the large allocation is senseless and
> would even if it worked, would likely need to be
> vmalloc/vfree
> >>>
> >>> Fair point!
> >>>
> >>> Pranith, Romanov, if you would like part of RCU that is less touchy
> >>> about random hacking, this would be a good place to start. Scripts in
> >>> tools/testing/selftests/rcutorture/bin do care about some of the format,
> >>> but the variable-length portion generated by cur_ops->stats() is as far
> >>> as I know only parsed by human eyes.
> >>>
> >>
> >> Here is a first run of the change. Please let me know if I am totally off.
> >> RFC. :)
> >
> > Thank you for taking this on!
>
> You are most welcome :)
>
> >
> >> Three things on Todo list:
> >>
> >> * We need to check that we are using less than the allocated size of the
> >> buffer (used > size). (we are allocating a big buffer, so not sure if
> >> necessary)
> >> * Need to check with the scripts if they are working.
> >> * I used a loop for pr_alert(). I am not sure this is right, there should
> >> be a better way for printing large buffers
> >>
> >> If the overall structure is ok I will go ahead and check how the scripts
> >> are handling these changes.
> >
> > One other thing... Convince this function (and a few others that it
> > calls) that the system really has 4096 CPUs, run this code, and see what
> > actually happens both before and after. Just to get a bit of practice
> > mixed in with the theory. ;-)
> >
>
> OK. I think to do this I need to use 4096 instead of nr_cpu_ids. I will try
> this and see how it goes :)
That's the spirit! You will probably have to adjust a few other things
as well.
Thanx, Paul
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Re: [RFC PATCH] use snprintf instead of sprintf in rcu_torture_printk
On 06/19/2014 07:49 PM, Paul E. McKenney wrote:
> On Thu, Jun 19, 2014 at 07:24:48PM -0400, Pranith Kumar wrote:
>> (dropping some CCs)
>>
>> On 06/19/2014 05:00 PM, Paul E. McKenney wrote:
>>> On Wed, Jun 18, 2014 at 12:49:42PM -0700, Joe Perches wrote:
I believe the function doesn't work well.
static void
rcu_torture_stats_print(void)
{
int size = nr_cpu_ids * 200 + 8192;
char *buf;
buf = kmalloc(size, GFP_KERNEL);
if (!buf) {
pr_err("rcu-torture: Out of memory, need: %d\n", size);
return;
}
rcu_torture_printk(buf);
pr_alert("%s", buf);
kfree(buf);
}
rcu_torture_printk simply fills buf
btw: I believe the arguments should pass size and
rcu_torture_printk should use snprintf/size
but all printks are limited to a maximum of 1024
bytes so the large allocation is senseless and
would even if it worked, would likely need to be
vmalloc/vfree
>>>
>>> Fair point!
>>>
>>> Pranith, Romanov, if you would like part of RCU that is less touchy
>>> about random hacking, this would be a good place to start. Scripts in
>>> tools/testing/selftests/rcutorture/bin do care about some of the format,
>>> but the variable-length portion generated by cur_ops->stats() is as far
>>> as I know only parsed by human eyes.
>>>
>>
>> Here is a first run of the change. Please let me know if I am totally off.
>> RFC. :)
>
> Thank you for taking this on!
You are most welcome :)
>
>> Three things on Todo list:
>>
>> * We need to check that we are using less than the allocated size of the
>> buffer (used > size). (we are allocating a big buffer, so not sure if
>> necessary)
>> * Need to check with the scripts if they are working.
>> * I used a loop for pr_alert(). I am not sure this is right, there should be
>> a better way for printing large buffers
>>
>> If the overall structure is ok I will go ahead and check how the scripts are
>> handling these changes.
>
> One other thing... Convince this function (and a few others that it
> calls) that the system really has 4096 CPUs, run this code, and see what
> actually happens both before and after. Just to get a bit of practice
> mixed in with the theory. ;-)
>
OK. I think to do this I need to use 4096 instead of nr_cpu_ids. I will try
this and see how it goes :)
--
Pranith
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Re: [RFC PATCH] use snprintf instead of sprintf in rcu_torture_printk
On 06/19/2014 07:33 PM, Joe Perches wrote:
>>
>> Here is a first run of the change. Please let me know if I am totally off.
>> RFC. :)
>>
>> Three things on Todo list:
>>
>> * We need to check that we are using less than the allocated size of the
>> buffer (used > size). (we are allocating a big buffer, so not sure if
>> necessary)
>> * Need to check with the scripts if they are working.
>> * I used a loop for pr_alert(). I am not sure this is right, there should be
>> a better way for printing large buffers
>>
>> If the overall structure is ok I will go ahead and check how the scripts are
>> handling these changes.
> []
>> diff --git a/kernel/rcu/rcutorture.c b/kernel/rcu/rcutorture.c
> []
>> @@ -1130,17 +1134,23 @@ rcu_torture_printk(char *page)
>> static void
>> rcu_torture_stats_print(void)
>> {
> []
>> +/* printk can print only print LOG_LINE_MAX chars at a time */
>> +while (used > 0) {
>> +size_pr = pr_alert("%s", buf);
>> +used -= size_pr;
>> +buf += size_pr;
>> +}
>> +vfree(buf);
>> }
>
> You can't increment buf then vfree it.
>
> Try a char *tmpbuf = buf, and increment it.
>
> I suggest you check by strcspn(tmpbuf, "\n")
> save the byte after, replace it with 0,
> print, replace the 0 with saved byte, and
> loop until end of buffer.
>
Please find v2 which I changed according to your comments:
v2:
* use tmpbuf to increment
* print each line (search for "\n" as suggested and mark with '\0')
* minor fix while calling cur_ops->stats
Signed-off-by: Pranith Kumar
---
include/linux/torture.h | 2 +-
kernel/rcu/rcutorture.c | 86 ++---
kernel/torture.c| 7 ++--
3 files changed, 57 insertions(+), 38 deletions(-)
diff --git a/include/linux/torture.h b/include/linux/torture.h
index 5ca58fc..1e47ec7 100644
--- a/include/linux/torture.h
+++ b/include/linux/torture.h
@@ -51,7 +51,7 @@
/* Definitions for online/offline exerciser. */
int torture_onoff_init(long ooholdoff, long oointerval);
-char *torture_onoff_stats(char *page);
+int torture_onoff_stats(char *page, int size);
bool torture_onoff_failures(void);
/* Low-rider random number generator. */
diff --git a/kernel/rcu/rcutorture.c b/kernel/rcu/rcutorture.c
index 7fa34f8..b149c5f 100644
--- a/kernel/rcu/rcutorture.c
+++ b/kernel/rcu/rcutorture.c
@@ -242,7 +242,7 @@ struct rcu_torture_ops {
void (*call)(struct rcu_head *head, void (*func)(struct rcu_head *rcu));
void (*cb_barrier)(void);
void (*fqs)(void);
- void (*stats)(char *page);
+ int (*stats)(char *page, int size);
int irq_capable;
int can_boost;
const char *name;
@@ -525,21 +525,22 @@ static void srcu_torture_barrier(void)
srcu_barrier(_ctl);
}
-static void srcu_torture_stats(char *page)
+static int srcu_torture_stats(char *page, int size)
{
- int cpu;
+ int cpu, used = 0;
int idx = srcu_ctl.completed & 0x1;
- page += sprintf(page, "%s%s per-CPU(idx=%d):",
+ used = snprintf(page, size, "%s%s per-CPU(idx=%d):",
torture_type, TORTURE_FLAG, idx);
for_each_possible_cpu(cpu) {
long c0, c1;
c0 = (long)per_cpu_ptr(srcu_ctl.per_cpu_ref, cpu)->c[!idx];
c1 = (long)per_cpu_ptr(srcu_ctl.per_cpu_ref, cpu)->c[idx];
- page += sprintf(page, " %d(%ld,%ld)", cpu, c0, c1);
+ used += snprintf(page + used, size - used, " %d(%ld,%ld)", cpu,
c0, c1);
}
- sprintf(page, "\n");
+ used += snprintf(page + used, size - used, "\n");
+ return used;
}
static void srcu_torture_synchronize_expedited(void)
@@ -1033,10 +1034,10 @@ rcu_torture_reader(void *arg)
/*
* Create an RCU-torture statistics message in the specified buffer.
*/
-static void
-rcu_torture_printk(char *page)
+static int
+rcu_torture_printk(char *page, int size)
{
- int cpu;
+ int cpu, used = 0;
int i;
long pipesummary[RCU_TORTURE_PIPE_LEN + 1] = { 0 };
long batchsummary[RCU_TORTURE_PIPE_LEN + 1] = { 0 };
@@ -1052,8 +1053,8 @@ rcu_torture_printk(char *page)
if (pipesummary[i] != 0)
break;
}
- page += sprintf(page, "%s%s ", torture_type, TORTURE_FLAG);
- page += sprintf(page,
+ used += snprintf(page + used, size - used, "%s%s ", torture_type,
TORTURE_FLAG);
+ used += snprintf(page + used, size - used,
"rtc: %p ver: %lu tfle: %d rta: %d rtaf: %d rtf: %d ",
rcu_torture_current,
rcu_torture_current_version,
@@ -1061,46 +1062,47 @@ rcu_torture_printk(char *page)
atomic_read(_rcu_torture_alloc),
atomic_read(_rcu_torture_alloc_fail),
atomic_read(_rcu_torture_free));
- page += sprintf(page, "rtmbe: %d rtbke: %ld rtbre: %ld ",
+ used +=
Re: [RFC PATCH] use snprintf instead of sprintf in rcu_torture_printk
On Thu, Jun 19, 2014 at 07:24:48PM -0400, Pranith Kumar wrote:
> (dropping some CCs)
>
> On 06/19/2014 05:00 PM, Paul E. McKenney wrote:
> > On Wed, Jun 18, 2014 at 12:49:42PM -0700, Joe Perches wrote:
> >>
> >> I believe the function doesn't work well.
> >>
> >> static void
> >> rcu_torture_stats_print(void)
> >> {
> >>int size = nr_cpu_ids * 200 + 8192;
> >>char *buf;
> >>
> >>buf = kmalloc(size, GFP_KERNEL);
> >>if (!buf) {
> >>pr_err("rcu-torture: Out of memory, need: %d\n", size);
> >>return;
> >>}
> >>rcu_torture_printk(buf);
> >>pr_alert("%s", buf);
> >>kfree(buf);
> >> }
> >>
> >> rcu_torture_printk simply fills buf
> >>
> >> btw: I believe the arguments should pass size and
> >> rcu_torture_printk should use snprintf/size
> >>
> >> but all printks are limited to a maximum of 1024
> >> bytes so the large allocation is senseless and
> >> would even if it worked, would likely need to be
> >> vmalloc/vfree
> >
> > Fair point!
> >
> > Pranith, Romanov, if you would like part of RCU that is less touchy
> > about random hacking, this would be a good place to start. Scripts in
> > tools/testing/selftests/rcutorture/bin do care about some of the format,
> > but the variable-length portion generated by cur_ops->stats() is as far
> > as I know only parsed by human eyes.
> >
>
> Here is a first run of the change. Please let me know if I am totally off.
> RFC. :)
Thank you for taking this on!
> Three things on Todo list:
>
> * We need to check that we are using less than the allocated size of the
> buffer (used > size). (we are allocating a big buffer, so not sure if
> necessary)
> * Need to check with the scripts if they are working.
> * I used a loop for pr_alert(). I am not sure this is right, there should be
> a better way for printing large buffers
>
> If the overall structure is ok I will go ahead and check how the scripts are
> handling these changes.
One other thing... Convince this function (and a few others that it
calls) that the system really has 4096 CPUs, run this code, and see what
actually happens both before and after. Just to get a bit of practice
mixed in with the theory. ;-)
Thanx, Paul
> >From 0d52fdcd941b0eaaacb6732f59f609595ac14d11 Mon Sep 17 00:00:00 2001
> From: Pranith Kumar
> Date: Thu, 19 Jun 2014 19:00:46 -0400
> Subject: [PATCH 1/1] use snprintf instead of sprintf in
> rcu_torture_print_stats
>
> Signed-off-by: Pranith Kumar
> ---
> include/linux/torture.h | 2 +-
> kernel/rcu/rcutorture.c | 76
> -
> kernel/torture.c| 7 +++--
> 3 files changed, 48 insertions(+), 37 deletions(-)
>
> diff --git a/include/linux/torture.h b/include/linux/torture.h
> index 5ca58fc..1e47ec7 100644
> --- a/include/linux/torture.h
> +++ b/include/linux/torture.h
> @@ -51,7 +51,7 @@
>
> /* Definitions for online/offline exerciser. */
> int torture_onoff_init(long ooholdoff, long oointerval);
> -char *torture_onoff_stats(char *page);
> +int torture_onoff_stats(char *page, int size);
> bool torture_onoff_failures(void);
>
> /* Low-rider random number generator. */
> diff --git a/kernel/rcu/rcutorture.c b/kernel/rcu/rcutorture.c
> index 7fa34f8..f708db4 100644
> --- a/kernel/rcu/rcutorture.c
> +++ b/kernel/rcu/rcutorture.c
> @@ -242,7 +242,7 @@ struct rcu_torture_ops {
> void (*call)(struct rcu_head *head, void (*func)(struct rcu_head *rcu));
> void (*cb_barrier)(void);
> void (*fqs)(void);
> - void (*stats)(char *page);
> + int (*stats)(char *page, int size);
> int irq_capable;
> int can_boost;
> const char *name;
> @@ -525,21 +525,22 @@ static void srcu_torture_barrier(void)
> srcu_barrier(_ctl);
> }
>
> -static void srcu_torture_stats(char *page)
> +static int srcu_torture_stats(char *page, int size)
> {
> - int cpu;
> + int cpu, used = 0;
> int idx = srcu_ctl.completed & 0x1;
>
> - page += sprintf(page, "%s%s per-CPU(idx=%d):",
> + used = snprintf(page, size, "%s%s per-CPU(idx=%d):",
> torture_type, TORTURE_FLAG, idx);
> for_each_possible_cpu(cpu) {
> long c0, c1;
>
> c0 = (long)per_cpu_ptr(srcu_ctl.per_cpu_ref, cpu)->c[!idx];
> c1 = (long)per_cpu_ptr(srcu_ctl.per_cpu_ref, cpu)->c[idx];
> - page += sprintf(page, " %d(%ld,%ld)", cpu, c0, c1);
> + used += snprintf(page + used, size - used, " %d(%ld,%ld)", cpu,
> c0, c1);
> }
> - sprintf(page, "\n");
> + used += snprintf(page + used, size - used, "\n");
> + return used;
> }
>
> static void srcu_torture_synchronize_expedited(void)
> @@ -1033,10 +1034,10 @@ rcu_torture_reader(void *arg)
> /*
> * Create an RCU-torture statistics message in the specified buffer.
> */
> -static void
> -rcu_torture_printk(char *page)
> +static int
> +rcu_torture_printk(char
Re: [RFC PATCH] use snprintf instead of sprintf in rcu_torture_printk
On Thu, 2014-06-19 at 19:24 -0400, Pranith Kumar wrote:
> (dropping some CCs)
>
> On 06/19/2014 05:00 PM, Paul E. McKenney wrote:
> > On Wed, Jun 18, 2014 at 12:49:42PM -0700, Joe Perches wrote:
> >>
> >> I believe the function doesn't work well.
> >>
> >> static void
> >> rcu_torture_stats_print(void)
> >> {
> >>int size = nr_cpu_ids * 200 + 8192;
> >>char *buf;
> >>
> >>buf = kmalloc(size, GFP_KERNEL);
> >>if (!buf) {
> >>pr_err("rcu-torture: Out of memory, need: %d\n", size);
> >>return;
> >>}
> >>rcu_torture_printk(buf);
> >>pr_alert("%s", buf);
> >>kfree(buf);
> >> }
> >>
> >> rcu_torture_printk simply fills buf
> >>
> >> btw: I believe the arguments should pass size and
> >> rcu_torture_printk should use snprintf/size
> >>
> >> but all printks are limited to a maximum of 1024
> >> bytes so the large allocation is senseless and
> >> would even if it worked, would likely need to be
> >> vmalloc/vfree
> >
> > Fair point!
> >
> > Pranith, Romanov, if you would like part of RCU that is less touchy
> > about random hacking, this would be a good place to start. Scripts in
> > tools/testing/selftests/rcutorture/bin do care about some of the format,
> > but the variable-length portion generated by cur_ops->stats() is as far
> > as I know only parsed by human eyes.
> >
>
> Here is a first run of the change. Please let me know if I am totally off.
> RFC. :)
>
> Three things on Todo list:
>
> * We need to check that we are using less than the allocated size of the
> buffer (used > size). (we are allocating a big buffer, so not sure if
> necessary)
> * Need to check with the scripts if they are working.
> * I used a loop for pr_alert(). I am not sure this is right, there should be
> a better way for printing large buffers
>
> If the overall structure is ok I will go ahead and check how the scripts are
> handling these changes.
[]
> diff --git a/kernel/rcu/rcutorture.c b/kernel/rcu/rcutorture.c
[]
> @@ -1130,17 +1134,23 @@ rcu_torture_printk(char *page)
> static void
> rcu_torture_stats_print(void)
> {
[]
> + /* printk can print only print LOG_LINE_MAX chars at a time */
> + while (used > 0) {
> + size_pr = pr_alert("%s", buf);
> + used -= size_pr;
> + buf += size_pr;
> + }
> + vfree(buf);
> }
You can't increment buf then vfree it.
Try a char *tmpbuf = buf, and increment it.
I suggest you check by strcspn(tmpbuf, "\n")
save the byte after, replace it with 0,
print, replace the 0 with saved byte, and
loop until end of buffer.
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
[RFC PATCH] use snprintf instead of sprintf in rcu_torture_printk
(dropping some CCs)
On 06/19/2014 05:00 PM, Paul E. McKenney wrote:
> On Wed, Jun 18, 2014 at 12:49:42PM -0700, Joe Perches wrote:
>>
>> I believe the function doesn't work well.
>>
>> static void
>> rcu_torture_stats_print(void)
>> {
>> int size = nr_cpu_ids * 200 + 8192;
>> char *buf;
>>
>> buf = kmalloc(size, GFP_KERNEL);
>> if (!buf) {
>> pr_err("rcu-torture: Out of memory, need: %d\n", size);
>> return;
>> }
>> rcu_torture_printk(buf);
>> pr_alert("%s", buf);
>> kfree(buf);
>> }
>>
>> rcu_torture_printk simply fills buf
>>
>> btw: I believe the arguments should pass size and
>> rcu_torture_printk should use snprintf/size
>>
>> but all printks are limited to a maximum of 1024
>> bytes so the large allocation is senseless and
>> would even if it worked, would likely need to be
>> vmalloc/vfree
>
> Fair point!
>
> Pranith, Romanov, if you would like part of RCU that is less touchy
> about random hacking, this would be a good place to start. Scripts in
> tools/testing/selftests/rcutorture/bin do care about some of the format,
> but the variable-length portion generated by cur_ops->stats() is as far
> as I know only parsed by human eyes.
>
Here is a first run of the change. Please let me know if I am totally off. RFC.
:)
Three things on Todo list:
* We need to check that we are using less than the allocated size of the buffer
(used > size). (we are allocating a big buffer, so not sure if necessary)
* Need to check with the scripts if they are working.
* I used a loop for pr_alert(). I am not sure this is right, there should be a
better way for printing large buffers
If the overall structure is ok I will go ahead and check how the scripts are
handling these changes.
>From 0d52fdcd941b0eaaacb6732f59f609595ac14d11 Mon Sep 17 00:00:00 2001
From: Pranith Kumar
Date: Thu, 19 Jun 2014 19:00:46 -0400
Subject: [PATCH 1/1] use snprintf instead of sprintf in rcu_torture_print_stats
Signed-off-by: Pranith Kumar
---
include/linux/torture.h | 2 +-
kernel/rcu/rcutorture.c | 76 -
kernel/torture.c| 7 +++--
3 files changed, 48 insertions(+), 37 deletions(-)
diff --git a/include/linux/torture.h b/include/linux/torture.h
index 5ca58fc..1e47ec7 100644
--- a/include/linux/torture.h
+++ b/include/linux/torture.h
@@ -51,7 +51,7 @@
/* Definitions for online/offline exerciser. */
int torture_onoff_init(long ooholdoff, long oointerval);
-char *torture_onoff_stats(char *page);
+int torture_onoff_stats(char *page, int size);
bool torture_onoff_failures(void);
/* Low-rider random number generator. */
diff --git a/kernel/rcu/rcutorture.c b/kernel/rcu/rcutorture.c
index 7fa34f8..f708db4 100644
--- a/kernel/rcu/rcutorture.c
+++ b/kernel/rcu/rcutorture.c
@@ -242,7 +242,7 @@ struct rcu_torture_ops {
void (*call)(struct rcu_head *head, void (*func)(struct rcu_head *rcu));
void (*cb_barrier)(void);
void (*fqs)(void);
- void (*stats)(char *page);
+ int (*stats)(char *page, int size);
int irq_capable;
int can_boost;
const char *name;
@@ -525,21 +525,22 @@ static void srcu_torture_barrier(void)
srcu_barrier(_ctl);
}
-static void srcu_torture_stats(char *page)
+static int srcu_torture_stats(char *page, int size)
{
- int cpu;
+ int cpu, used = 0;
int idx = srcu_ctl.completed & 0x1;
- page += sprintf(page, "%s%s per-CPU(idx=%d):",
+ used = snprintf(page, size, "%s%s per-CPU(idx=%d):",
torture_type, TORTURE_FLAG, idx);
for_each_possible_cpu(cpu) {
long c0, c1;
c0 = (long)per_cpu_ptr(srcu_ctl.per_cpu_ref, cpu)->c[!idx];
c1 = (long)per_cpu_ptr(srcu_ctl.per_cpu_ref, cpu)->c[idx];
- page += sprintf(page, " %d(%ld,%ld)", cpu, c0, c1);
+ used += snprintf(page + used, size - used, " %d(%ld,%ld)", cpu,
c0, c1);
}
- sprintf(page, "\n");
+ used += snprintf(page + used, size - used, "\n");
+ return used;
}
static void srcu_torture_synchronize_expedited(void)
@@ -1033,10 +1034,10 @@ rcu_torture_reader(void *arg)
/*
* Create an RCU-torture statistics message in the specified buffer.
*/
-static void
-rcu_torture_printk(char *page)
+static int
+rcu_torture_printk(char *page, int size)
{
- int cpu;
+ int cpu, used = 0;
int i;
long pipesummary[RCU_TORTURE_PIPE_LEN + 1] = { 0 };
long batchsummary[RCU_TORTURE_PIPE_LEN + 1] = { 0 };
@@ -1052,8 +1053,8 @@ rcu_torture_printk(char *page)
if (pipesummary[i] != 0)
break;
}
- page += sprintf(page, "%s%s ", torture_type, TORTURE_FLAG);
- page += sprintf(page,
+ used += snprintf(page + used, size - used, "%s%s ", torture_type,
TORTURE_FLAG);
+ used += snprintf(page + used, size - used,
"rtc:
[RFC PATCH] use snprintf instead of sprintf in rcu_torture_printk
(dropping some CCs)
On 06/19/2014 05:00 PM, Paul E. McKenney wrote:
On Wed, Jun 18, 2014 at 12:49:42PM -0700, Joe Perches wrote:
I believe the function doesn't work well.
static void
rcu_torture_stats_print(void)
{
int size = nr_cpu_ids * 200 + 8192;
char *buf;
buf = kmalloc(size, GFP_KERNEL);
if (!buf) {
pr_err(rcu-torture: Out of memory, need: %d\n, size);
return;
}
rcu_torture_printk(buf);
pr_alert(%s, buf);
kfree(buf);
}
rcu_torture_printk simply fills buf
btw: I believe the arguments should pass size and
rcu_torture_printk should use snprintf/size
but all printks are limited to a maximum of 1024
bytes so the large allocation is senseless and
would even if it worked, would likely need to be
vmalloc/vfree
Fair point!
Pranith, Romanov, if you would like part of RCU that is less touchy
about random hacking, this would be a good place to start. Scripts in
tools/testing/selftests/rcutorture/bin do care about some of the format,
but the variable-length portion generated by cur_ops-stats() is as far
as I know only parsed by human eyes.
Here is a first run of the change. Please let me know if I am totally off. RFC.
:)
Three things on Todo list:
* We need to check that we are using less than the allocated size of the buffer
(used size). (we are allocating a big buffer, so not sure if necessary)
* Need to check with the scripts if they are working.
* I used a loop for pr_alert(). I am not sure this is right, there should be a
better way for printing large buffers
If the overall structure is ok I will go ahead and check how the scripts are
handling these changes.
From 0d52fdcd941b0eaaacb6732f59f609595ac14d11 Mon Sep 17 00:00:00 2001
From: Pranith Kumar bobby.pr...@gmail.com
Date: Thu, 19 Jun 2014 19:00:46 -0400
Subject: [PATCH 1/1] use snprintf instead of sprintf in rcu_torture_print_stats
Signed-off-by: Pranith Kumar bobby.pr...@gmail.com
---
include/linux/torture.h | 2 +-
kernel/rcu/rcutorture.c | 76 -
kernel/torture.c| 7 +++--
3 files changed, 48 insertions(+), 37 deletions(-)
diff --git a/include/linux/torture.h b/include/linux/torture.h
index 5ca58fc..1e47ec7 100644
--- a/include/linux/torture.h
+++ b/include/linux/torture.h
@@ -51,7 +51,7 @@
/* Definitions for online/offline exerciser. */
int torture_onoff_init(long ooholdoff, long oointerval);
-char *torture_onoff_stats(char *page);
+int torture_onoff_stats(char *page, int size);
bool torture_onoff_failures(void);
/* Low-rider random number generator. */
diff --git a/kernel/rcu/rcutorture.c b/kernel/rcu/rcutorture.c
index 7fa34f8..f708db4 100644
--- a/kernel/rcu/rcutorture.c
+++ b/kernel/rcu/rcutorture.c
@@ -242,7 +242,7 @@ struct rcu_torture_ops {
void (*call)(struct rcu_head *head, void (*func)(struct rcu_head *rcu));
void (*cb_barrier)(void);
void (*fqs)(void);
- void (*stats)(char *page);
+ int (*stats)(char *page, int size);
int irq_capable;
int can_boost;
const char *name;
@@ -525,21 +525,22 @@ static void srcu_torture_barrier(void)
srcu_barrier(srcu_ctl);
}
-static void srcu_torture_stats(char *page)
+static int srcu_torture_stats(char *page, int size)
{
- int cpu;
+ int cpu, used = 0;
int idx = srcu_ctl.completed 0x1;
- page += sprintf(page, %s%s per-CPU(idx=%d):,
+ used = snprintf(page, size, %s%s per-CPU(idx=%d):,
torture_type, TORTURE_FLAG, idx);
for_each_possible_cpu(cpu) {
long c0, c1;
c0 = (long)per_cpu_ptr(srcu_ctl.per_cpu_ref, cpu)-c[!idx];
c1 = (long)per_cpu_ptr(srcu_ctl.per_cpu_ref, cpu)-c[idx];
- page += sprintf(page, %d(%ld,%ld), cpu, c0, c1);
+ used += snprintf(page + used, size - used, %d(%ld,%ld), cpu,
c0, c1);
}
- sprintf(page, \n);
+ used += snprintf(page + used, size - used, \n);
+ return used;
}
static void srcu_torture_synchronize_expedited(void)
@@ -1033,10 +1034,10 @@ rcu_torture_reader(void *arg)
/*
* Create an RCU-torture statistics message in the specified buffer.
*/
-static void
-rcu_torture_printk(char *page)
+static int
+rcu_torture_printk(char *page, int size)
{
- int cpu;
+ int cpu, used = 0;
int i;
long pipesummary[RCU_TORTURE_PIPE_LEN + 1] = { 0 };
long batchsummary[RCU_TORTURE_PIPE_LEN + 1] = { 0 };
@@ -1052,8 +1053,8 @@ rcu_torture_printk(char *page)
if (pipesummary[i] != 0)
break;
}
- page += sprintf(page, %s%s , torture_type, TORTURE_FLAG);
- page += sprintf(page,
+ used += snprintf(page + used, size - used, %s%s , torture_type,
TORTURE_FLAG);
+ used += snprintf(page + used, size - used,
rtc: %p ver: %lu tfle: %d rta: %d rtaf: %d rtf: %d ,
Re: [RFC PATCH] use snprintf instead of sprintf in rcu_torture_printk
On Thu, 2014-06-19 at 19:24 -0400, Pranith Kumar wrote:
(dropping some CCs)
On 06/19/2014 05:00 PM, Paul E. McKenney wrote:
On Wed, Jun 18, 2014 at 12:49:42PM -0700, Joe Perches wrote:
I believe the function doesn't work well.
static void
rcu_torture_stats_print(void)
{
int size = nr_cpu_ids * 200 + 8192;
char *buf;
buf = kmalloc(size, GFP_KERNEL);
if (!buf) {
pr_err(rcu-torture: Out of memory, need: %d\n, size);
return;
}
rcu_torture_printk(buf);
pr_alert(%s, buf);
kfree(buf);
}
rcu_torture_printk simply fills buf
btw: I believe the arguments should pass size and
rcu_torture_printk should use snprintf/size
but all printks are limited to a maximum of 1024
bytes so the large allocation is senseless and
would even if it worked, would likely need to be
vmalloc/vfree
Fair point!
Pranith, Romanov, if you would like part of RCU that is less touchy
about random hacking, this would be a good place to start. Scripts in
tools/testing/selftests/rcutorture/bin do care about some of the format,
but the variable-length portion generated by cur_ops-stats() is as far
as I know only parsed by human eyes.
Here is a first run of the change. Please let me know if I am totally off.
RFC. :)
Three things on Todo list:
* We need to check that we are using less than the allocated size of the
buffer (used size). (we are allocating a big buffer, so not sure if
necessary)
* Need to check with the scripts if they are working.
* I used a loop for pr_alert(). I am not sure this is right, there should be
a better way for printing large buffers
If the overall structure is ok I will go ahead and check how the scripts are
handling these changes.
[]
diff --git a/kernel/rcu/rcutorture.c b/kernel/rcu/rcutorture.c
[]
@@ -1130,17 +1134,23 @@ rcu_torture_printk(char *page)
static void
rcu_torture_stats_print(void)
{
[]
+ /* printk can print only print LOG_LINE_MAX chars at a time */
+ while (used 0) {
+ size_pr = pr_alert(%s, buf);
+ used -= size_pr;
+ buf += size_pr;
+ }
+ vfree(buf);
}
You can't increment buf then vfree it.
Try a char *tmpbuf = buf, and increment it.
I suggest you check by strcspn(tmpbuf, \n)
save the byte after, replace it with 0,
print, replace the 0 with saved byte, and
loop until end of buffer.
--
To unsubscribe from this list: send the line unsubscribe linux-kernel in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Re: [RFC PATCH] use snprintf instead of sprintf in rcu_torture_printk
On Thu, Jun 19, 2014 at 07:24:48PM -0400, Pranith Kumar wrote:
(dropping some CCs)
On 06/19/2014 05:00 PM, Paul E. McKenney wrote:
On Wed, Jun 18, 2014 at 12:49:42PM -0700, Joe Perches wrote:
I believe the function doesn't work well.
static void
rcu_torture_stats_print(void)
{
int size = nr_cpu_ids * 200 + 8192;
char *buf;
buf = kmalloc(size, GFP_KERNEL);
if (!buf) {
pr_err(rcu-torture: Out of memory, need: %d\n, size);
return;
}
rcu_torture_printk(buf);
pr_alert(%s, buf);
kfree(buf);
}
rcu_torture_printk simply fills buf
btw: I believe the arguments should pass size and
rcu_torture_printk should use snprintf/size
but all printks are limited to a maximum of 1024
bytes so the large allocation is senseless and
would even if it worked, would likely need to be
vmalloc/vfree
Fair point!
Pranith, Romanov, if you would like part of RCU that is less touchy
about random hacking, this would be a good place to start. Scripts in
tools/testing/selftests/rcutorture/bin do care about some of the format,
but the variable-length portion generated by cur_ops-stats() is as far
as I know only parsed by human eyes.
Here is a first run of the change. Please let me know if I am totally off.
RFC. :)
Thank you for taking this on!
Three things on Todo list:
* We need to check that we are using less than the allocated size of the
buffer (used size). (we are allocating a big buffer, so not sure if
necessary)
* Need to check with the scripts if they are working.
* I used a loop for pr_alert(). I am not sure this is right, there should be
a better way for printing large buffers
If the overall structure is ok I will go ahead and check how the scripts are
handling these changes.
One other thing... Convince this function (and a few others that it
calls) that the system really has 4096 CPUs, run this code, and see what
actually happens both before and after. Just to get a bit of practice
mixed in with the theory. ;-)
Thanx, Paul
From 0d52fdcd941b0eaaacb6732f59f609595ac14d11 Mon Sep 17 00:00:00 2001
From: Pranith Kumar bobby.pr...@gmail.com
Date: Thu, 19 Jun 2014 19:00:46 -0400
Subject: [PATCH 1/1] use snprintf instead of sprintf in
rcu_torture_print_stats
Signed-off-by: Pranith Kumar bobby.pr...@gmail.com
---
include/linux/torture.h | 2 +-
kernel/rcu/rcutorture.c | 76
-
kernel/torture.c| 7 +++--
3 files changed, 48 insertions(+), 37 deletions(-)
diff --git a/include/linux/torture.h b/include/linux/torture.h
index 5ca58fc..1e47ec7 100644
--- a/include/linux/torture.h
+++ b/include/linux/torture.h
@@ -51,7 +51,7 @@
/* Definitions for online/offline exerciser. */
int torture_onoff_init(long ooholdoff, long oointerval);
-char *torture_onoff_stats(char *page);
+int torture_onoff_stats(char *page, int size);
bool torture_onoff_failures(void);
/* Low-rider random number generator. */
diff --git a/kernel/rcu/rcutorture.c b/kernel/rcu/rcutorture.c
index 7fa34f8..f708db4 100644
--- a/kernel/rcu/rcutorture.c
+++ b/kernel/rcu/rcutorture.c
@@ -242,7 +242,7 @@ struct rcu_torture_ops {
void (*call)(struct rcu_head *head, void (*func)(struct rcu_head *rcu));
void (*cb_barrier)(void);
void (*fqs)(void);
- void (*stats)(char *page);
+ int (*stats)(char *page, int size);
int irq_capable;
int can_boost;
const char *name;
@@ -525,21 +525,22 @@ static void srcu_torture_barrier(void)
srcu_barrier(srcu_ctl);
}
-static void srcu_torture_stats(char *page)
+static int srcu_torture_stats(char *page, int size)
{
- int cpu;
+ int cpu, used = 0;
int idx = srcu_ctl.completed 0x1;
- page += sprintf(page, %s%s per-CPU(idx=%d):,
+ used = snprintf(page, size, %s%s per-CPU(idx=%d):,
torture_type, TORTURE_FLAG, idx);
for_each_possible_cpu(cpu) {
long c0, c1;
c0 = (long)per_cpu_ptr(srcu_ctl.per_cpu_ref, cpu)-c[!idx];
c1 = (long)per_cpu_ptr(srcu_ctl.per_cpu_ref, cpu)-c[idx];
- page += sprintf(page, %d(%ld,%ld), cpu, c0, c1);
+ used += snprintf(page + used, size - used, %d(%ld,%ld), cpu,
c0, c1);
}
- sprintf(page, \n);
+ used += snprintf(page + used, size - used, \n);
+ return used;
}
static void srcu_torture_synchronize_expedited(void)
@@ -1033,10 +1034,10 @@ rcu_torture_reader(void *arg)
/*
* Create an RCU-torture statistics message in the specified buffer.
*/
-static void
-rcu_torture_printk(char *page)
+static int
+rcu_torture_printk(char *page, int size)
{
- int cpu;
+ int cpu, used = 0;
int i;
long pipesummary[RCU_TORTURE_PIPE_LEN + 1] = { 0 };
long
Re: [RFC PATCH] use snprintf instead of sprintf in rcu_torture_printk
On 06/19/2014 07:33 PM, Joe Perches wrote:
Here is a first run of the change. Please let me know if I am totally off.
RFC. :)
Three things on Todo list:
* We need to check that we are using less than the allocated size of the
buffer (used size). (we are allocating a big buffer, so not sure if
necessary)
* Need to check with the scripts if they are working.
* I used a loop for pr_alert(). I am not sure this is right, there should be
a better way for printing large buffers
If the overall structure is ok I will go ahead and check how the scripts are
handling these changes.
[]
diff --git a/kernel/rcu/rcutorture.c b/kernel/rcu/rcutorture.c
[]
@@ -1130,17 +1134,23 @@ rcu_torture_printk(char *page)
static void
rcu_torture_stats_print(void)
{
[]
+/* printk can print only print LOG_LINE_MAX chars at a time */
+while (used 0) {
+size_pr = pr_alert(%s, buf);
+used -= size_pr;
+buf += size_pr;
+}
+vfree(buf);
}
You can't increment buf then vfree it.
Try a char *tmpbuf = buf, and increment it.
I suggest you check by strcspn(tmpbuf, \n)
save the byte after, replace it with 0,
print, replace the 0 with saved byte, and
loop until end of buffer.
Please find v2 which I changed according to your comments:
v2:
* use tmpbuf to increment
* print each line (search for \n as suggested and mark with '\0')
* minor fix while calling cur_ops-stats
Signed-off-by: Pranith Kumar bobby.pr...@gmail.com
---
include/linux/torture.h | 2 +-
kernel/rcu/rcutorture.c | 86 ++---
kernel/torture.c| 7 ++--
3 files changed, 57 insertions(+), 38 deletions(-)
diff --git a/include/linux/torture.h b/include/linux/torture.h
index 5ca58fc..1e47ec7 100644
--- a/include/linux/torture.h
+++ b/include/linux/torture.h
@@ -51,7 +51,7 @@
/* Definitions for online/offline exerciser. */
int torture_onoff_init(long ooholdoff, long oointerval);
-char *torture_onoff_stats(char *page);
+int torture_onoff_stats(char *page, int size);
bool torture_onoff_failures(void);
/* Low-rider random number generator. */
diff --git a/kernel/rcu/rcutorture.c b/kernel/rcu/rcutorture.c
index 7fa34f8..b149c5f 100644
--- a/kernel/rcu/rcutorture.c
+++ b/kernel/rcu/rcutorture.c
@@ -242,7 +242,7 @@ struct rcu_torture_ops {
void (*call)(struct rcu_head *head, void (*func)(struct rcu_head *rcu));
void (*cb_barrier)(void);
void (*fqs)(void);
- void (*stats)(char *page);
+ int (*stats)(char *page, int size);
int irq_capable;
int can_boost;
const char *name;
@@ -525,21 +525,22 @@ static void srcu_torture_barrier(void)
srcu_barrier(srcu_ctl);
}
-static void srcu_torture_stats(char *page)
+static int srcu_torture_stats(char *page, int size)
{
- int cpu;
+ int cpu, used = 0;
int idx = srcu_ctl.completed 0x1;
- page += sprintf(page, %s%s per-CPU(idx=%d):,
+ used = snprintf(page, size, %s%s per-CPU(idx=%d):,
torture_type, TORTURE_FLAG, idx);
for_each_possible_cpu(cpu) {
long c0, c1;
c0 = (long)per_cpu_ptr(srcu_ctl.per_cpu_ref, cpu)-c[!idx];
c1 = (long)per_cpu_ptr(srcu_ctl.per_cpu_ref, cpu)-c[idx];
- page += sprintf(page, %d(%ld,%ld), cpu, c0, c1);
+ used += snprintf(page + used, size - used, %d(%ld,%ld), cpu,
c0, c1);
}
- sprintf(page, \n);
+ used += snprintf(page + used, size - used, \n);
+ return used;
}
static void srcu_torture_synchronize_expedited(void)
@@ -1033,10 +1034,10 @@ rcu_torture_reader(void *arg)
/*
* Create an RCU-torture statistics message in the specified buffer.
*/
-static void
-rcu_torture_printk(char *page)
+static int
+rcu_torture_printk(char *page, int size)
{
- int cpu;
+ int cpu, used = 0;
int i;
long pipesummary[RCU_TORTURE_PIPE_LEN + 1] = { 0 };
long batchsummary[RCU_TORTURE_PIPE_LEN + 1] = { 0 };
@@ -1052,8 +1053,8 @@ rcu_torture_printk(char *page)
if (pipesummary[i] != 0)
break;
}
- page += sprintf(page, %s%s , torture_type, TORTURE_FLAG);
- page += sprintf(page,
+ used += snprintf(page + used, size - used, %s%s , torture_type,
TORTURE_FLAG);
+ used += snprintf(page + used, size - used,
rtc: %p ver: %lu tfle: %d rta: %d rtaf: %d rtf: %d ,
rcu_torture_current,
rcu_torture_current_version,
@@ -1061,46 +1062,47 @@ rcu_torture_printk(char *page)
atomic_read(n_rcu_torture_alloc),
atomic_read(n_rcu_torture_alloc_fail),
atomic_read(n_rcu_torture_free));
- page += sprintf(page, rtmbe: %d rtbke: %ld rtbre: %ld ,
+ used += snprintf(page + used, size - used, rtmbe: %d rtbke: %ld rtbre:
%ld ,
Re: [RFC PATCH] use snprintf instead of sprintf in rcu_torture_printk
On 06/19/2014 07:49 PM, Paul E. McKenney wrote:
On Thu, Jun 19, 2014 at 07:24:48PM -0400, Pranith Kumar wrote:
(dropping some CCs)
On 06/19/2014 05:00 PM, Paul E. McKenney wrote:
On Wed, Jun 18, 2014 at 12:49:42PM -0700, Joe Perches wrote:
I believe the function doesn't work well.
static void
rcu_torture_stats_print(void)
{
int size = nr_cpu_ids * 200 + 8192;
char *buf;
buf = kmalloc(size, GFP_KERNEL);
if (!buf) {
pr_err(rcu-torture: Out of memory, need: %d\n, size);
return;
}
rcu_torture_printk(buf);
pr_alert(%s, buf);
kfree(buf);
}
rcu_torture_printk simply fills buf
btw: I believe the arguments should pass size and
rcu_torture_printk should use snprintf/size
but all printks are limited to a maximum of 1024
bytes so the large allocation is senseless and
would even if it worked, would likely need to be
vmalloc/vfree
Fair point!
Pranith, Romanov, if you would like part of RCU that is less touchy
about random hacking, this would be a good place to start. Scripts in
tools/testing/selftests/rcutorture/bin do care about some of the format,
but the variable-length portion generated by cur_ops-stats() is as far
as I know only parsed by human eyes.
Here is a first run of the change. Please let me know if I am totally off.
RFC. :)
Thank you for taking this on!
You are most welcome :)
Three things on Todo list:
* We need to check that we are using less than the allocated size of the
buffer (used size). (we are allocating a big buffer, so not sure if
necessary)
* Need to check with the scripts if they are working.
* I used a loop for pr_alert(). I am not sure this is right, there should be
a better way for printing large buffers
If the overall structure is ok I will go ahead and check how the scripts are
handling these changes.
One other thing... Convince this function (and a few others that it
calls) that the system really has 4096 CPUs, run this code, and see what
actually happens both before and after. Just to get a bit of practice
mixed in with the theory. ;-)
OK. I think to do this I need to use 4096 instead of nr_cpu_ids. I will try
this and see how it goes :)
--
Pranith
--
To unsubscribe from this list: send the line unsubscribe linux-kernel in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Re: [RFC PATCH] use snprintf instead of sprintf in rcu_torture_printk
On Thu, Jun 19, 2014 at 08:13:22PM -0400, Pranith Kumar wrote:
On 06/19/2014 07:49 PM, Paul E. McKenney wrote:
On Thu, Jun 19, 2014 at 07:24:48PM -0400, Pranith Kumar wrote:
(dropping some CCs)
On 06/19/2014 05:00 PM, Paul E. McKenney wrote:
On Wed, Jun 18, 2014 at 12:49:42PM -0700, Joe Perches wrote:
I believe the function doesn't work well.
static void
rcu_torture_stats_print(void)
{
int size = nr_cpu_ids * 200 + 8192;
char *buf;
buf = kmalloc(size, GFP_KERNEL);
if (!buf) {
pr_err(rcu-torture: Out of memory, need: %d\n, size);
return;
}
rcu_torture_printk(buf);
pr_alert(%s, buf);
kfree(buf);
}
rcu_torture_printk simply fills buf
btw: I believe the arguments should pass size and
rcu_torture_printk should use snprintf/size
but all printks are limited to a maximum of 1024
bytes so the large allocation is senseless and
would even if it worked, would likely need to be
vmalloc/vfree
Fair point!
Pranith, Romanov, if you would like part of RCU that is less touchy
about random hacking, this would be a good place to start. Scripts in
tools/testing/selftests/rcutorture/bin do care about some of the format,
but the variable-length portion generated by cur_ops-stats() is as far
as I know only parsed by human eyes.
Here is a first run of the change. Please let me know if I am totally off.
RFC. :)
Thank you for taking this on!
You are most welcome :)
Three things on Todo list:
* We need to check that we are using less than the allocated size of the
buffer (used size). (we are allocating a big buffer, so not sure if
necessary)
* Need to check with the scripts if they are working.
* I used a loop for pr_alert(). I am not sure this is right, there should
be a better way for printing large buffers
If the overall structure is ok I will go ahead and check how the scripts
are handling these changes.
One other thing... Convince this function (and a few others that it
calls) that the system really has 4096 CPUs, run this code, and see what
actually happens both before and after. Just to get a bit of practice
mixed in with the theory. ;-)
OK. I think to do this I need to use 4096 instead of nr_cpu_ids. I will try
this and see how it goes :)
That's the spirit! You will probably have to adjust a few other things
as well.
Thanx, Paul
--
To unsubscribe from this list: send the line unsubscribe linux-kernel in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
