speakup-r empty line lockup

[Okash Khawaja]

Hi,

The lockup when running speakup-r at start of an empty line occurs
because simulated key press is generated (using speakup_fake_down_arrow)
from context of keyboard_notifier_call callback which is called in
interrupt context. The simulated keypress leads to
keybaord_notifier_call to be trigerred again from the same context
leading to the lockup. The exact cause could be priority inversion where
simulated keypress cannot be processed because there is a real keyboard
interrupt already being processed (the one from which simulated keypress
was triggered), hence causing a deadlock. Please share your thoughts on
this. Here is the call chain.

(speakup-r) --> keyboard_notifier_call --> speakup_key --> do_spkup -->
read_all_doc --> get_sentence_buf [which returns -1 because of empty
line] --> kbd_fakekey2(RA_DOWN_ARROW) --> speakup_fake_down_arrow

The following patch resolves this by not simulating the keypress inside
keyboard notifier callback but instead delegating it to cursor_timer. In
the above chain, when get_sentence_buf returns -1, this patch starts
timer and passes RA_DOWN_ARROW as argument. When timer handler runs and
sees RA_DOWN_ARROW, it will then call kbd_fakekey2(RA_DOWN_ARROW) which
will correctly simulate the keypress inside timer context. I've tested
this succesfully.

It's the first time I've worked on this side of the code so please
review carefully :)

Finally, I have updated the test repo on github so you can test from
there.

Thanks,
Okash

---
drivers/staging/speakup/main.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)

--- a/drivers/staging/speakup/main.c
+++ b/drivers/staging/speakup/main.c
@@ -1408,7 +1408,8 @@ static void read_all_doc(struct vc_data
cursor_track = read_all_mode;
spk_reset_index_count(0);
if (get_sentence_buf(vc, 0) == -1) {
- kbd_fakekey2(vc, RA_DOWN_ARROW);
+ del_timer(&cursor_timer);
+ start_read_all_timer(vc, RA_DOWN_ARROW);
} else {
say_sentence_num(0, 0);
synth_insert_next_index(0);

[Samuel Thibault]

Hello,

Post by Okash Khawaja
It's the first time I've worked on this side of the code so please
review carefully :)

Well, I'd say you're on your own on this one, unless elder developers
remember something :)

From what I can see and understand, I'd say

Reviewed-by: Samuel Thibault <***@ens-lyon.org>

Samuel

[Chris Brannon]

Post by Okash Khawaja
The following patch resolves this by not simulating the keypress inside
keyboard notifier callback but instead delegating it to cursor_timer. In
the above chain, when get_sentence_buf returns -1, this patch starts
timer and passes RA_DOWN_ARROW as argument. When timer handler runs and
sees RA_DOWN_ARROW, it will then call kbd_fakekey2(RA_DOWN_ARROW) which
will correctly simulate the keypress inside timer context. I've tested
this succesfully.

I wrote the fakekey code back in 2010, but I don't remember details.
Your reasoning looks very, very sound, so you get my tentative

Reviewed-by: Chris Brannon <***@the-brannons.com>

[John Covici]

Thanks a lot, I will definitely test this -- do I need a new
export.patch as well?

On Sat, 24 Jun 2017 05:06:45 -0400,

Post by Okash Khawaja
Hi,
The lockup when running speakup-r at start of an empty line occurs
because simulated key press is generated (using speakup_fake_down_arrow)
from context of keyboard_notifier_call callback which is called in
interrupt context. The simulated keypress leads to
keybaord_notifier_call to be trigerred again from the same context
leading to the lockup. The exact cause could be priority inversion where
simulated keypress cannot be processed because there is a real keyboard
interrupt already being processed (the one from which simulated keypress
was triggered), hence causing a deadlock. Please share your thoughts on
this. Here is the call chain.
(speakup-r) --> keyboard_notifier_call --> speakup_key --> do_spkup -->
read_all_doc --> get_sentence_buf [which returns -1 because of empty
line] --> kbd_fakekey2(RA_DOWN_ARROW) --> speakup_fake_down_arrow
The following patch resolves this by not simulating the keypress inside
keyboard notifier callback but instead delegating it to cursor_timer. In
the above chain, when get_sentence_buf returns -1, this patch starts
timer and passes RA_DOWN_ARROW as argument. When timer handler runs and
sees RA_DOWN_ARROW, it will then call kbd_fakekey2(RA_DOWN_ARROW) which
will correctly simulate the keypress inside timer context. I've tested
this succesfully.
It's the first time I've worked on this side of the code so please
review carefully :)
Finally, I have updated the test repo on github so you can test from
there.
Thanks,
Okash
---
drivers/staging/speakup/main.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
--- a/drivers/staging/speakup/main.c
+++ b/drivers/staging/speakup/main.c
@@ -1408,7 +1408,8 @@ static void read_all_doc(struct vc_data
cursor_track = read_all_mode;
spk_reset_index_count(0);
if (get_sentence_buf(vc, 0) == -1) {
- kbd_fakekey2(vc, RA_DOWN_ARROW);
+ del_timer(&cursor_timer);
+ start_read_all_timer(vc, RA_DOWN_ARROW);
} else {
say_sentence_num(0, 0);
synth_insert_next_index(0);

[Okash Khawaja]

Yes export-tty.patch on the repo has been updated. Let me know if you get any problems.

Cheers,
Okash

Post by John Covici
Thanks a lot, I will definitely test this -- do I need a new
export.patch as well?
On Sat, 24 Jun 2017 05:06:45 -0400,

Post by Okash Khawaja
Hi,
The lockup when running speakup-r at start of an empty line occurs
because simulated key press is generated (using speakup_fake_down_arrow)
from context of keyboard_notifier_call callback which is called in
interrupt context. The simulated keypress leads to
keybaord_notifier_call to be trigerred again from the same context
leading to the lockup. The exact cause could be priority inversion where
simulated keypress cannot be processed because there is a real keyboard
interrupt already being processed (the one from which simulated keypress
was triggered), hence causing a deadlock. Please share your thoughts on
this. Here is the call chain.
(speakup-r) --> keyboard_notifier_call --> speakup_key --> do_spkup -->
read_all_doc --> get_sentence_buf [which returns -1 because of empty
line] --> kbd_fakekey2(RA_DOWN_ARROW) --> speakup_fake_down_arrow
The following patch resolves this by not simulating the keypress inside
keyboard notifier callback but instead delegating it to cursor_timer. In
the above chain, when get_sentence_buf returns -1, this patch starts
timer and passes RA_DOWN_ARROW as argument. When timer handler runs and
sees RA_DOWN_ARROW, it will then call kbd_fakekey2(RA_DOWN_ARROW) which
will correctly simulate the keypress inside timer context. I've tested
this succesfully.
It's the first time I've worked on this side of the code so please
review carefully :)
Finally, I have updated the test repo on github so you can test from
there.
Thanks,
Okash
---
drivers/staging/speakup/main.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
--- a/drivers/staging/speakup/main.c
+++ b/drivers/staging/speakup/main.c
@@ -1408,7 +1408,8 @@ static void read_all_doc(struct vc_data
cursor_track = read_all_mode;
spk_reset_index_count(0);
if (get_sentence_buf(vc, 0) == -1) {
- kbd_fakekey2(vc, RA_DOWN_ARROW);
+ del_timer(&cursor_timer);
+ start_read_all_timer(vc, RA_DOWN_ARROW);
} else {
say_sentence_num(0, 0);
synth_insert_next_index(0);

--
How do
you spend it?
John Covici

[John Covici]

I just tested the empty line lockup is gone, but now speakup-r does
not work, actually in the file I had, after reading a number of lines
when I hit control, the cursor was actually at the end of the file.

Thanks for working on this one.

On Sat, 24 Jun 2017 05:06:45 -0400,

Post by Okash Khawaja
Hi,
The lockup when running speakup-r at start of an empty line occurs
because simulated key press is generated (using speakup_fake_down_arrow)
from context of keyboard_notifier_call callback which is called in
interrupt context. The simulated keypress leads to
keybaord_notifier_call to be trigerred again from the same context
leading to the lockup. The exact cause could be priority inversion where
simulated keypress cannot be processed because there is a real keyboard
interrupt already being processed (the one from which simulated keypress
was triggered), hence causing a deadlock. Please share your thoughts on
this. Here is the call chain.
(speakup-r) --> keyboard_notifier_call --> speakup_key --> do_spkup -->
read_all_doc --> get_sentence_buf [which returns -1 because of empty
line] --> kbd_fakekey2(RA_DOWN_ARROW) --> speakup_fake_down_arrow
The following patch resolves this by not simulating the keypress inside
keyboard notifier callback but instead delegating it to cursor_timer. In
the above chain, when get_sentence_buf returns -1, this patch starts
timer and passes RA_DOWN_ARROW as argument. When timer handler runs and
sees RA_DOWN_ARROW, it will then call kbd_fakekey2(RA_DOWN_ARROW) which
will correctly simulate the keypress inside timer context. I've tested
this succesfully.
It's the first time I've worked on this side of the code so please
review carefully :)
Finally, I have updated the test repo on github so you can test from
there.
Thanks,
Okash
---
drivers/staging/speakup/main.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
--- a/drivers/staging/speakup/main.c
+++ b/drivers/staging/speakup/main.c
@@ -1408,7 +1408,8 @@ static void read_all_doc(struct vc_data
cursor_track = read_all_mode;
spk_reset_index_count(0);
if (get_sentence_buf(vc, 0) == -1) {
- kbd_fakekey2(vc, RA_DOWN_ARROW);
+ del_timer(&cursor_timer);
+ start_read_all_timer(vc, RA_DOWN_ARROW);
} else {
say_sentence_num(0, 0);
synth_insert_next_index(0);

[Okash Khawaja]

Hi,

This looks like the problem you raised previously, the one which was
resolved by calling flush_buffers. Is this happening on ttyUSB?
usb-serial driver doesn't implement flush_buffers method which means
that speakup-r will have this issue until there is flush_buffers in
usb-serial.

Thanks,
Okash

Post by John Covici
I just tested the empty line lockup is gone, but now speakup-r does
not work, actually in the file I had, after reading a number of lines
when I hit control, the cursor was actually at the end of the file.
Thanks for working on this one.
On Sat, 24 Jun 2017 05:06:45 -0400,

Post by Okash Khawaja
Hi,
The lockup when running speakup-r at start of an empty line occurs
because simulated key press is generated (using speakup_fake_down_arrow)
from context of keyboard_notifier_call callback which is called in
interrupt context. The simulated keypress leads to
keybaord_notifier_call to be trigerred again from the same context
leading to the lockup. The exact cause could be priority inversion where
simulated keypress cannot be processed because there is a real keyboard
interrupt already being processed (the one from which simulated keypress
was triggered), hence causing a deadlock. Please share your thoughts on
this. Here is the call chain.
(speakup-r) --> keyboard_notifier_call --> speakup_key --> do_spkup -->
read_all_doc --> get_sentence_buf [which returns -1 because of empty
line] --> kbd_fakekey2(RA_DOWN_ARROW) --> speakup_fake_down_arrow
The following patch resolves this by not simulating the keypress inside
keyboard notifier callback but instead delegating it to cursor_timer. In
the above chain, when get_sentence_buf returns -1, this patch starts
timer and passes RA_DOWN_ARROW as argument. When timer handler runs and
sees RA_DOWN_ARROW, it will then call kbd_fakekey2(RA_DOWN_ARROW) which
will correctly simulate the keypress inside timer context. I've tested
this succesfully.
It's the first time I've worked on this side of the code so please
review carefully :)
Finally, I have updated the test repo on github so you can test from
there.
Thanks,
Okash
---
drivers/staging/speakup/main.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
--- a/drivers/staging/speakup/main.c
+++ b/drivers/staging/speakup/main.c
@@ -1408,7 +1408,8 @@ static void read_all_doc(struct vc_data
cursor_track = read_all_mode;
spk_reset_index_count(0);
if (get_sentence_buf(vc, 0) == -1) {
- kbd_fakekey2(vc, RA_DOWN_ARROW);
+ del_timer(&cursor_timer);
+ start_read_all_timer(vc, RA_DOWN_ARROW);
} else {
say_sentence_num(0, 0);
synth_insert_next_index(0);

--
How do
you spend it?
John Covici

[Okash Khawaja]

When I test speakup-r with speakup_soft + espeakup, it doesn't behave
the way it should. That is true with or without this patch. May be I've
missed something but here is what I did.

Open spkguide.txt in vim. Move cursor to beginning of the line
"Permission is granted to copy...". Hit keyboard ins + r. Cursor
immediately moves to beginning of next line "under the terms of..." and
the speaking starts. Speaking stops after finishing the next line, i.e.
the "under the terms of..." line. It stops automatically. So at this
stage cursor is at beginning of that line while speaking stopped at end
of that line.

This is behaviour is same both with and without the patch, so I am
unable to test it.

Again, if this problem happens on ttyUSB* but doesn't happen on ttyS*
then that is because usb-serial doesn't implement flush data.

Thanks,
Okash

Post by Okash Khawaja
Hi,
This looks like the problem you raised previously, the one which was
resolved by calling flush_buffers. Is this happening on ttyUSB?
usb-serial driver doesn't implement flush_buffers method which means
that speakup-r will have this issue until there is flush_buffers in
usb-serial.
Thanks,
Okash

Post by John Covici
I just tested the empty line lockup is gone, but now speakup-r does
not work, actually in the file I had, after reading a number of lines
when I hit control, the cursor was actually at the end of the file.
Thanks for working on this one.
On Sat, 24 Jun 2017 05:06:45 -0400,

Post by Okash Khawaja
Hi,
The lockup when running speakup-r at start of an empty line occurs
because simulated key press is generated (using speakup_fake_down_arrow)
from context of keyboard_notifier_call callback which is called in
interrupt context. The simulated keypress leads to
keybaord_notifier_call to be trigerred again from the same context
leading to the lockup. The exact cause could be priority inversion where
simulated keypress cannot be processed because there is a real keyboard
interrupt already being processed (the one from which simulated keypress
was triggered), hence causing a deadlock. Please share your thoughts on
this. Here is the call chain.
(speakup-r) --> keyboard_notifier_call --> speakup_key --> do_spkup -->
read_all_doc --> get_sentence_buf [which returns -1 because of empty
line] --> kbd_fakekey2(RA_DOWN_ARROW) --> speakup_fake_down_arrow
The following patch resolves this by not simulating the keypress inside
keyboard notifier callback but instead delegating it to cursor_timer. In
the above chain, when get_sentence_buf returns -1, this patch starts
timer and passes RA_DOWN_ARROW as argument. When timer handler runs and
sees RA_DOWN_ARROW, it will then call kbd_fakekey2(RA_DOWN_ARROW) which
will correctly simulate the keypress inside timer context. I've tested
this succesfully.
It's the first time I've worked on this side of the code so please
review carefully :)
Finally, I have updated the test repo on github so you can test from
there.
Thanks,
Okash
---
drivers/staging/speakup/main.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
--- a/drivers/staging/speakup/main.c
+++ b/drivers/staging/speakup/main.c
@@ -1408,7 +1408,8 @@ static void read_all_doc(struct vc_data
cursor_track = read_all_mode;
spk_reset_index_count(0);
if (get_sentence_buf(vc, 0) == -1) {
- kbd_fakekey2(vc, RA_DOWN_ARROW);
+ del_timer(&cursor_timer);
+ start_read_all_timer(vc, RA_DOWN_ARROW);
} else {
say_sentence_num(0, 0);
synth_insert_next_index(0);

--
How do
you spend it?
John Covici

[John Covici]

Mope, its on regular serial, somehow your fix broke it.

On Mon, 03 Jul 2017 07:14:53 -0400,

Post by Okash Khawaja
Hi,
This looks like the problem you raised previously, the one which was
resolved by calling flush_buffers. Is this happening on ttyUSB?
usb-serial driver doesn't implement flush_buffers method which means
that speakup-r will have this issue until there is flush_buffers in
usb-serial.
Thanks,
Okash

Post by John Covici
I just tested the empty line lockup is gone, but now speakup-r does
not work, actually in the file I had, after reading a number of lines
when I hit control, the cursor was actually at the end of the file.
Thanks for working on this one.
On Sat, 24 Jun 2017 05:06:45 -0400,

Post by Okash Khawaja
Hi,
The lockup when running speakup-r at start of an empty line occurs
because simulated key press is generated (using speakup_fake_down_arrow)
from context of keyboard_notifier_call callback which is called in
interrupt context. The simulated keypress leads to
keybaord_notifier_call to be trigerred again from the same context
leading to the lockup. The exact cause could be priority inversion where
simulated keypress cannot be processed because there is a real keyboard
interrupt already being processed (the one from which simulated keypress
was triggered), hence causing a deadlock. Please share your thoughts on
this. Here is the call chain.
(speakup-r) --> keyboard_notifier_call --> speakup_key --> do_spkup -->
read_all_doc --> get_sentence_buf [which returns -1 because of empty
line] --> kbd_fakekey2(RA_DOWN_ARROW) --> speakup_fake_down_arrow
The following patch resolves this by not simulating the keypress inside
keyboard notifier callback but instead delegating it to cursor_timer. In
the above chain, when get_sentence_buf returns -1, this patch starts
timer and passes RA_DOWN_ARROW as argument. When timer handler runs and
sees RA_DOWN_ARROW, it will then call kbd_fakekey2(RA_DOWN_ARROW) which
will correctly simulate the keypress inside timer context. I've tested
this succesfully.
It's the first time I've worked on this side of the code so please
review carefully :)
Finally, I have updated the test repo on github so you can test from
there.
Thanks,
Okash
---
drivers/staging/speakup/main.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
--- a/drivers/staging/speakup/main.c
+++ b/drivers/staging/speakup/main.c
@@ -1408,7 +1408,8 @@ static void read_all_doc(struct vc_data
cursor_track = read_all_mode;
spk_reset_index_count(0);
if (get_sentence_buf(vc, 0) == -1) {
- kbd_fakekey2(vc, RA_DOWN_ARROW);
+ del_timer(&cursor_timer);
+ start_read_all_timer(vc, RA_DOWN_ARROW);
} else {
say_sentence_num(0, 0);
synth_insert_next_index(0);

--
How do
you spend it?
John Covici

[Okash Khawaja]

Thanks, I'm looking into it.

Post by John Covici
Mope, its on regular serial, somehow your fix broke it.
On Mon, 03 Jul 2017 07:14:53 -0400,

Post by Okash Khawaja
Hi,
This looks like the problem you raised previously, the one which was
resolved by calling flush_buffers. Is this happening on ttyUSB?
usb-serial driver doesn't implement flush_buffers method which means
that speakup-r will have this issue until there is flush_buffers in
usb-serial.
Thanks,
Okash

Post by John Covici
I just tested the empty line lockup is gone, but now speakup-r does
not work, actually in the file I had, after reading a number of lines
when I hit control, the cursor was actually at the end of the file.
Thanks for working on this one.
On Sat, 24 Jun 2017 05:06:45 -0400,

Post by Okash Khawaja
Hi,
The lockup when running speakup-r at start of an empty line occurs
because simulated key press is generated (using speakup_fake_down_arrow)
from context of keyboard_notifier_call callback which is called in
interrupt context. The simulated keypress leads to
keybaord_notifier_call to be trigerred again from the same context
leading to the lockup. The exact cause could be priority inversion where
simulated keypress cannot be processed because there is a real keyboard
interrupt already being processed (the one from which simulated keypress
was triggered), hence causing a deadlock. Please share your thoughts on
this. Here is the call chain.
(speakup-r) --> keyboard_notifier_call --> speakup_key --> do_spkup -->
read_all_doc --> get_sentence_buf [which returns -1 because of empty
line] --> kbd_fakekey2(RA_DOWN_ARROW) --> speakup_fake_down_arrow
The following patch resolves this by not simulating the keypress inside
keyboard notifier callback but instead delegating it to cursor_timer. In
the above chain, when get_sentence_buf returns -1, this patch starts
timer and passes RA_DOWN_ARROW as argument. When timer handler runs and
sees RA_DOWN_ARROW, it will then call kbd_fakekey2(RA_DOWN_ARROW) which
will correctly simulate the keypress inside timer context. I've tested
this succesfully.
It's the first time I've worked on this side of the code so please
review carefully :)
Finally, I have updated the test repo on github so you can test from
there.
Thanks,
Okash
---
drivers/staging/speakup/main.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
--- a/drivers/staging/speakup/main.c
+++ b/drivers/staging/speakup/main.c
@@ -1408,7 +1408,8 @@ static void read_all_doc(struct vc_data
cursor_track = read_all_mode;
spk_reset_index_count(0);
if (get_sentence_buf(vc, 0) == -1) {
- kbd_fakekey2(vc, RA_DOWN_ARROW);
+ del_timer(&cursor_timer);
+ start_read_all_timer(vc, RA_DOWN_ARROW);
} else {
say_sentence_num(0, 0);
synth_insert_next_index(0);

--
How do
you spend it?
John Covici

--
How do
you spend it?
John Covici

[Okash Khawaja]

Hi,

I used printk to output result of in_interrupt() when the affected code
is called. It looks like the affected code is only called when inside
interrupt context, i.e. when keyboard_notifier_callback is called and
that's when we can't call speakup_fake_down_arrow as that leads to
lockup, as explained earlier in this thread.

Given limited scope of the change - it is only called when running
speakup-r from empty line - I can't work out how this results in cursor
going to the end of file when control is hit. However, I say this based
on speakup_soft + espeakup test in which speakup-r stops automatically
after reading two lines (both before and after this patch). So until I
am able to replicate the problem, it will be hard for me to test. The
triple talk I bricked, I haven't got round to fixing it yet - if it can
be fixed.

If there is any additional information that might be relevant, that will
be helpful. Until then, I will try to work out what might be happening.

Since the github repo contains this and other changes, it will be
interesting to check if the problem is down to this change or is a
result of other changes that also went in.

Thanks,
Okash

Post by John Covici
Mope, its on regular serial, somehow your fix broke it.
On Mon, 03 Jul 2017 07:14:53 -0400,

Post by Okash Khawaja
Hi,
This looks like the problem you raised previously, the one which was
resolved by calling flush_buffers. Is this happening on ttyUSB?
usb-serial driver doesn't implement flush_buffers method which means
that speakup-r will have this issue until there is flush_buffers in
usb-serial.
Thanks,
Okash

Post by John Covici
I just tested the empty line lockup is gone, but now speakup-r does
not work, actually in the file I had, after reading a number of lines
when I hit control, the cursor was actually at the end of the file.
Thanks for working on this one.
On Sat, 24 Jun 2017 05:06:45 -0400,

Post by Okash Khawaja
Hi,
The lockup when running speakup-r at start of an empty line occurs
because simulated key press is generated (using speakup_fake_down_arrow)
from context of keyboard_notifier_call callback which is called in
interrupt context. The simulated keypress leads to
keybaord_notifier_call to be trigerred again from the same context
leading to the lockup. The exact cause could be priority inversion where
simulated keypress cannot be processed because there is a real keyboard
interrupt already being processed (the one from which simulated keypress
was triggered), hence causing a deadlock. Please share your thoughts on
this. Here is the call chain.
(speakup-r) --> keyboard_notifier_call --> speakup_key --> do_spkup -->
read_all_doc --> get_sentence_buf [which returns -1 because of empty
line] --> kbd_fakekey2(RA_DOWN_ARROW) --> speakup_fake_down_arrow
The following patch resolves this by not simulating the keypress inside
keyboard notifier callback but instead delegating it to cursor_timer. In
the above chain, when get_sentence_buf returns -1, this patch starts
timer and passes RA_DOWN_ARROW as argument. When timer handler runs and
sees RA_DOWN_ARROW, it will then call kbd_fakekey2(RA_DOWN_ARROW) which
will correctly simulate the keypress inside timer context. I've tested
this succesfully.
It's the first time I've worked on this side of the code so please
review carefully :)
Finally, I have updated the test repo on github so you can test from
there.
Thanks,
Okash
---
drivers/staging/speakup/main.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
--- a/drivers/staging/speakup/main.c
+++ b/drivers/staging/speakup/main.c
@@ -1408,7 +1408,8 @@ static void read_all_doc(struct vc_data
cursor_track = read_all_mode;
spk_reset_index_count(0);
if (get_sentence_buf(vc, 0) == -1) {
- kbd_fakekey2(vc, RA_DOWN_ARROW);
+ del_timer(&cursor_timer);
+ start_read_all_timer(vc, RA_DOWN_ARROW);
} else {
say_sentence_num(0, 0);
synth_insert_next_index(0);

--
How do
you spend it?
John Covici

--
How do
you spend it?
John Covici

[Chris Brannon]

Post by Okash Khawaja
Given limited scope of the change - it is only called when running
speakup-r from empty line - I can't work out how this results in cursor
going to the end of file when control is hit. However, I say this based
on speakup_soft + espeakup test in which speakup-r stops automatically
after reading two lines (both before and after this patch).

espeakup was never modified to support speakup-r. I think speechd-up
may have done it. Code is at https://github.com/WilliamH/speechd-up but
I don't know what is involved in getting it running these days.

-- Chris

[John Covici]

speakup-r did work with speechd-up, last time I tried it.

On Wed, 05 Jul 2017 16:00:05 -0400,

Post by Chris Brannon

Post by Okash Khawaja
Given limited scope of the change - it is only called when running
speakup-r from empty line - I can't work out how this results in cursor
going to the end of file when control is hit. However, I say this based
on speakup_soft + espeakup test in which speakup-r stops automatically
after reading two lines (both before and after this patch).

espeakup was never modified to support speakup-r. I think speechd-up
may have done it. Code is at https://github.com/WilliamH/speechd-up but
I don't know what is involved in getting it running these days.
-- Chris

[Okash Khawaja]

Awesome. That's great

Post by John Covici
speakup-r did work with speechd-up, last time I tried it.
On Wed, 05 Jul 2017 16:00:05 -0400,

Post by Chris Brannon

Post by Okash Khawaja
Given limited scope of the change - it is only called when running
speakup-r from empty line - I can't work out how this results in cursor
going to the end of file when control is hit. However, I say this based
on speakup_soft + espeakup test in which speakup-r stops automatically
after reading two lines (both before and after this patch).

espeakup was never modified to support speakup-r. I think speechd-up
may have done it. Code is at https://github.com/WilliamH/speechd-up but
I don't know what is involved in getting it running these days.
-- Chris

--
How do
you spend it?
John Covici

[Okash Khawaja]

Hi John,

I'm now trying to replicate the problem where cursor goes to end of file
when hitting control after reading several lines with speakup-r. I'm
testing it on spkguide.txt. Starting from beginning I went up to line 45.
Hitting control stopped reading and cursor was at beginning of next line to
the one being read.

Would you suggest reading more lines? BTW, when running speakup-r, cursor
always jumps to beginning of next line while reading current line. That's
same, with or without the change I made.

Thanks,
Okash

Post by Okash Khawaja
Awesome. That's great

Post by John Covici
speakup-r did work with speechd-up, last time I tried it.
On Wed, 05 Jul 2017 16:00:05 -0400,

Post by Chris Brannon

Post by Okash Khawaja
Given limited scope of the change - it is only called when running
speakup-r from empty line - I can't work out how this results in cursor
going to the end of file when control is hit. However, I say this based
on speakup_soft + espeakup test in which speakup-r stops automatically
after reading two lines (both before and after this patch).

espeakup was never modified to support speakup-r. I think speechd-up
may have done it. Code is at https://github.com/WilliamH/speechd-up

but

Post by John Covici

Post by Chris Brannon
I don't know what is involved in getting it running these days.
-- Chris

--
How do
you spend it?
John Covici

[John Covici]

hmmm, I will have to try this again. I know I didn't read that many
lines. I think it went to the end because it ran out of lines. If
you say its not doing that for you, I will have to retest and see what
happens.

On Fri, 18 Aug 2017 03:17:33 -0400,

[1 <text/plain; UTF-8 (7bit)>]
[2 <text/html; UTF-8 (quoted-printable)>]
Hi John,
I'm now trying to replicate the problem where cursor goes to end of file when hitting control after reading several lines with speakup-r. I'm testing it on spkguide.txt. Starting from beginning I went up to line 45. Hitting control
stopped reading and cursor was at beginning of next line to the one being read.
Would you suggest reading more lines? BTW, when running speakup-r, cursor always jumps to beginning of next line while reading current line. That's same, with or without the change I made.
Thanks,
Okash
Awesome. That's great

Post by John Covici
speakup-r did work with speechd-up, last time I tried it.
On Wed, 05 Jul 2017 16:00:05 -0400,

Post by Okash Khawaja
Given limited scope of the change - it is only called when running
speakup-r from empty line - I can't work out how this results in cursor
going to the end of file when control is hit. However, I say this based
on speakup_soft + espeakup test in which speakup-r stops automatically
after reading two lines (both before and after this patch).

espeakup was never modified to support speakup-r. I think speechd-up
may have done it. Code is at https://github.com/WilliamH/speechd-up but
I don't know what is involved in getting it running these days.
-- Chris

--
How do
you spend it?
John Covici

[Samuel Thibault]

Post by Okash Khawaja
Would you suggest reading more lines?

The point of the feature is that by typing control, you stop the speech
at the place where you want to bring the cursor to :)

Samuel

[Okash Khawaja]

Post by Samuel Thibault

Post by Okash Khawaja
Would you suggest reading more lines?

The point of the feature is that by typing control, you stop the speech
at the place where you want to bring the cursor to :)

I see... So the use case is that you can move the cursor to a line
where you want the speech to stop and hit control, so that hitting
speakup-r again will then start speaking from that line. Is that right?

Thanks,
Okash

[Kirk Reiser]

Hahahah, close. The object is you can hit speakup-r and speakup will
start reading the document from where the cursor is until you tap the
control, shift or whatever silence key and it will place the cursor at
the position you heard whent you tapped the key. That way one can
continuously read a document in an edittor and have the cursor able to
start again where you hit the silence key for further reading or
editting.

Kirk

Post by Okash Khawaja

Post by Samuel Thibault

Post by Okash Khawaja
Would you suggest reading more lines?

The point of the feature is that by typing control, you stop the speech
at the place where you want to bring the cursor to :)

I see... So the use case is that you can move the cursor to a line
where you want the speech to stop and hit control, so that hitting
speakup-r again will then start speaking from that line. Is that right?
Thanks,
Okash
_______________________________________________
Speakup mailing list
http://linux-speakup.org/cgi-bin/mailman/listinfo/speakup

--
Well that's it then, colour me secure!

-----BEGIN PGP PUBLIC KEY BLOCK-----
Version: GnuPG v1
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=nrPH
-----END PGP PUBLIC KEY BLOCK-----

[Chime Hart]

Well Kirk-and-All, I sure wish that feature would work while reading a live mail
in pine
Chime

[Janina Sajka]

This will be absolutely awesome, especially if can persist it's even as
you might use Speakup in another terminal/console for some other task
before resuming.

So, is it particular to a tty? Or, should we not expect to run away to
another tty for some quick task before resuming reading?

Just trying to get my mind around a cool feature I expect to use a lot!

Janina

Post by Kirk Reiser
Hahahah, close. The object is you can hit speakup-r and speakup will
start reading the document from where the cursor is until you tap the
control, shift or whatever silence key and it will place the cursor at
the position you heard whent you tapped the key. That way one can
continuously read a document in an edittor and have the cursor able to
start again where you hit the silence key for further reading or
editting.
Kirk

Post by Okash Khawaja

Post by Samuel Thibault

Post by Okash Khawaja
Would you suggest reading more lines?

The point of the feature is that by typing control, you stop the speech
at the place where you want to bring the cursor to :)

I see... So the use case is that you can move the cursor to a line
where you want the speech to stop and hit control, so that hitting
speakup-r again will then start speaking from that line. Is that right?
Thanks,
Okash
_______________________________________________
Speakup mailing list
http://linux-speakup.org/cgi-bin/mailman/listinfo/speakup

--
Well that's it then, colour me secure!
-----BEGIN PGP PUBLIC KEY BLOCK-----
Version: GnuPG v1
mQENBFYV5DMBCAC060mbsnLhGPjnFkf0R0p+7MxcfxlOuy5wc8y59y9ZNF0RZD1s
OTEsDih4vD9YJ3zA78VsBUDK47aiDWduh3nHzYN2ZSuxAQ9u7qPqphCG0jPagTU8
p7+Ceeya4I5odWtq+Nkf1UrHB7KKEtexphStSwUG5Bhi4bb84YinmX/a3I+OGV1D
by4QBSdPvSuDw0qFkt/ucLyEwv4L6lDjoH2GF+tnCew4SJtliJFvA1k7NpWO6HW9
aWtBxfYU85ccZKBSE25y+9KprUCncVTpaVs3FztCWG0dQRXHvEbV+Damp/IBd9Jv
HZX7azqbERUa/FjPTIlZhhI9VtaZaFfJSH+5ABEBAAG0HUtpcmsgUmVpc2VyIDxr
aXJrQHJlaXNlcnMuY2E+iQE+BBMBAgAoBQJWFeQzAhsDBQkADS8ABgsJCAcDAgYV
CAIJCgsEFgIDAQIeAQIXgAAKCRAHTEsk7UQUUoeuB/wIqsdLCfDrSvr3qg7rKBDg
ru44OMuRit6hbdWFZjmxccCdjeNhBJRVd5wrEqjj5YoqQAhmacXaEB0DO/TZlDgo
kUfJM7lrtQD4mYU9GVtrzJxCJoBUyeMVnMJt39F91tBu0mYM6oI/dv81dwxIv++4
hj55TZ4GG7DGYAy4LwNb+noNbivgOFHlnfNq8nxhZkHbJdYKP+sptZOL5sagmBQZ
iS9STB54g/U7Jtt1Fe+JwDmbxQhbSHa9JuWn0xZ8CtYhrz06xSqZl5vpMlak3eW2
x6m6IcqZfyuI2K7W/9BCgcsQyYzufO4Gk9KyPNISskX6pFBLuNxIH6hdfxSYYm9y
uQENBFYV5DMBCACtMyhHog5MR6eQUPTx7fWH5ntkgCtmWvQp4lcKj0HHbteDWglS
NVbWKWEk9PAKA4UeQVUH4vOhTRhAPpuDUavLdp2tDtT7ZBVh91B3AWIM6+7fIvyU
2uYt1q/CNjga8RllXBT7mW2zHGEYQFIkBJvqlU0PN1HlxRZIbSSEb+zQuVAd+ph3
kt/oZon3ZbNmKg+arsYMmKkYJ0REwKQib7h5Xl31aK74XmWBp2Ky+lopsJSP8wpH
AfC71h4s3LDm8ADHF1Ns4KuGZdLTugr8uiPm5kEJFGes1uYKy8R7OTFko0NEuJkv
STfpPYnTU2qDCJBH08zZErI/6YBIlSsCSde3ABEBAAGJASUEGAECAA8FAlYV5DMC
GwwFCQANLwAACgkQB0xLJO1EFFKAmgf/d3dk1/HgmF8rmvYVru/hJvmIpmiLqPl5
bYSwdZeU+k82qp3xACM2yMJhOh89SgHsaaqQAE1qo5rAJcSG7/+7M/kzf4u/WM/E
unXDtLkbzi5Zl+gjoikrfOhgF0NmuGdlrOme8a6ue7+iE4XLAo0/jhVlh45O6Iq0
0DGyeFr22cR3jZj4wRmPw5zj4r/sWc06UfquVAEMmfIvJMaGYvwBI+TU6gI8MjLe
VDY0vay/nQ79fXSLQmYEvjwKXIavQu9c8TFt0z9EDdoIMx69ZunqZuYQInxaT+cL
i9zhihMGz4XA1q3blLNX3I0jWzAa23ZchI7htc3kfxp1jWqrGyGEIg==
=nrPH
-----END PGP PUBLIC KEY BLOCK-----
_______________________________________________
Speakup mailing list
http://linux-speakup.org/cgi-bin/mailman/listinfo/speakup

--
Janina Sajka, Phone: [phone]
sip:***@asterisk.rednote.net
Email: ***@rednote.net

Linux Foundation Fellow
Executive Chair, Accessibility Workgroup: http://a11y.org

The World Wide Web Consortium (W3C), Web Accessibility Initiative (WAI)
Chair, Accessible Platform Architectures http://www.w3.org/wai/apa

[John Covici]

While speakup-r is going, you need to stay in that console. this is
the one thing I wish eemacspeak had. It works great except for that
empty line lockup. David Borowski's mods fixed that problem, so if
you don't mind using the old serial code, that will work as well --
and if you use speakup with speech dispatcher and speechd-up speakup-r
will work as well -- not with espeakup. I use this every day, would
not have a speech system without it.


On Mon, 21 Aug 2017 10:49:20 -0400,

Post by Janina Sajka
This will be absolutely awesome, especially if can persist it's even as
you might use Speakup in another terminal/console for some other task
before resuming.
So, is it particular to a tty? Or, should we not expect to run away to
another tty for some quick task before resuming reading?
Just trying to get my mind around a cool feature I expect to use a lot!
Janina

Post by Kirk Reiser
Hahahah, close. The object is you can hit speakup-r and speakup will
start reading the document from where the cursor is until you tap the
control, shift or whatever silence key and it will place the cursor at
the position you heard whent you tapped the key. That way one can
continuously read a document in an edittor and have the cursor able to
start again where you hit the silence key for further reading or
editting.
Kirk

Post by Okash Khawaja

Post by Samuel Thibault

Post by Okash Khawaja
Would you suggest reading more lines?

The point of the feature is that by typing control, you stop the speech
at the place where you want to bring the cursor to :)

I see... So the use case is that you can move the cursor to a line
where you want the speech to stop and hit control, so that hitting
speakup-r again will then start speaking from that line. Is that right?
Thanks,
Okash
_______________________________________________
Speakup mailing list
http://linux-speakup.org/cgi-bin/mailman/listinfo/speakup

--
Well that's it then, colour me secure!
-----BEGIN PGP PUBLIC KEY BLOCK-----
Version: GnuPG v1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=nrPH
-----END PGP PUBLIC KEY BLOCK-----
_______________________________________________
Speakup mailing list
http://linux-speakup.org/cgi-bin/mailman/listinfo/speakup

--
Janina Sajka, Phone: [phone]
Linux Foundation Fellow
Executive Chair, Accessibility Workgroup: http://a11y.org
The World Wide Web Consortium (W3C), Web Accessibility Initiative (WAI)
Chair, Accessible Platform Architectures http://www.w3.org/wai/apa
_______________________________________________
Speakup mailing list
http://linux-speakup.org/cgi-bin/mailman/listinfo/speakup

[Samuel Thibault]

It works great except for that empty line lockup. David Borowski's
mods fixed that problem,

Is it known what exactly he did to fix it? That would probably very
informative.

Samuel

[John Covici]

He rewrote the whole part of the code, adding some new features as
well, so the bug mysteriously went away.

On Mon, 21 Aug 2017 12:05:25 -0400,

Post by Samuel Thibault

It works great except for that empty line lockup. David Borowski's
mods fixed that problem,

Is it known what exactly he did to fix it? That would probably very
informative.
Samuel

[Okash Khawaja]

Hi,

Post by John Covici
He rewrote the whole part of the code, adding some new features as
well, so the bug mysteriously went away.

I still think the root cause is correct, i.e. interrupt being triggered
from its own interrupt handler context. The fix however was too broad.
First one changed the behaviour in all cases. Second one from last night
narrowed the change to just the interrupt context.

If I had right serial synth, I would try the attached patch. This
narrows the fix down even further, so that we avoid calling
speakup_fake_down_arrow() only when we are in context of
keyboard_notifier_call() which is always invoked in interrupt context,
in response to a keyboard event.

Thanks,
Okash

[John Covici]

OK, tested the patch, no joy with this one either. Note that a couple
of thehunks sucedded with a bit of line difference, but that should
not be of significance. The procedure I used was to unpatch the
previous one and then put the new one in, same like I did with the
previous patch, unpatch old one and put in new one.

On Mon, 21 Aug 2017 15:24:32 -0400,

[1 <text/plain; us-ascii (7bit)>]
Hi,

Post by John Covici
He rewrote the whole part of the code, adding some new features as
well, so the bug mysteriously went away.

I still think the root cause is correct, i.e. interrupt being triggered
from its own interrupt handler context. The fix however was too broad.
First one changed the behaviour in all cases. Second one from last night
narrowed the change to just the interrupt context.
If I had right serial synth, I would try the attached patch. This
narrows the fix down even further, so that we avoid calling
speakup_fake_down_arrow() only when we are in context of
keyboard_notifier_call() which is always invoked in interrupt context,
in response to a keyboard event.
Thanks,
Okash
[2 speakup-r-fix <text/plain; us-ascii (7bit)>]
---
drivers/staging/speakup/main.c | 15 ++++++++++++---
1 file changed, 12 insertions(+), 3 deletions(-)
--- a/drivers/staging/speakup/main.c
+++ b/drivers/staging/speakup/main.c
@@ -1376,6 +1376,8 @@ static void reset_highlight_buffers(stru
static int read_all_key;
+static volatile int in_keyboard_notifier = 0;
+
static void start_read_all_timer(struct vc_data *vc, int command);
enum {
@@ -1408,7 +1410,10 @@ static void read_all_doc(struct vc_data
cursor_track = read_all_mode;
spk_reset_index_count(0);
if (get_sentence_buf(vc, 0) == -1) {
- kbd_fakekey2(vc, RA_DOWN_ARROW);
+ del_timer(&cursor_timer);
+ if (!in_keyboard_notifier)
+ speakup_fake_down_arrow();
+ start_read_all_timer(vc, RA_DOWN_ARROW);
} else {
say_sentence_num(0, 0);
synth_insert_next_index(0);
@@ -2212,8 +2217,10 @@ static int keyboard_notifier_call(struct
int ret = NOTIFY_OK;
static int keycode; /* to hold the current keycode */
+ in_keyboard_notifier = 1;
+
if (vc->vc_mode == KD_GRAPHICS)
- return ret;
+ goto out;
/*
* First, determine whether we are handling a fake keypress on
@@ -2225,7 +2232,7 @@ static int keyboard_notifier_call(struct
*/
if (speakup_fake_key_pressed())
- return ret;
+ goto out;
switch (code) {
@@ -2266,6 +2273,8 @@ static int keyboard_notifier_call(struct
break;
}
}
+ in_keyboard_notifier = 0;
return ret;
}

[Okash Khawaja]

Thanks John. Appreciate it.

I'll update here if I make progress.

Cheers,
Okash

Post by John Covici
OK, tested the patch, no joy with this one either. Note that a couple
of thehunks sucedded with a bit of line difference, but that should
not be of significance. The procedure I used was to unpatch the
previous one and then put the new one in, same like I did with the
previous patch, unpatch old one and put in new one.
On Mon, 21 Aug 2017 15:24:32 -0400,

[1 <text/plain; us-ascii (7bit)>]
Hi,

Post by John Covici
He rewrote the whole part of the code, adding some new features as
well, so the bug mysteriously went away.

I still think the root cause is correct, i.e. interrupt being triggered
from its own interrupt handler context. The fix however was too broad.
First one changed the behaviour in all cases. Second one from last night
narrowed the change to just the interrupt context.
If I had right serial synth, I would try the attached patch. This
narrows the fix down even further, so that we avoid calling
speakup_fake_down_arrow() only when we are in context of
keyboard_notifier_call() which is always invoked in interrupt context,
in response to a keyboard event.
Thanks,
Okash
[2 speakup-r-fix <text/plain; us-ascii (7bit)>]
---
drivers/staging/speakup/main.c | 15 ++++++++++++---
1 file changed, 12 insertions(+), 3 deletions(-)
--- a/drivers/staging/speakup/main.c
+++ b/drivers/staging/speakup/main.c
@@ -1376,6 +1376,8 @@ static void reset_highlight_buffers(stru
static int read_all_key;
+static volatile int in_keyboard_notifier = 0;
+
static void start_read_all_timer(struct vc_data *vc, int command);
enum {
@@ -1408,7 +1410,10 @@ static void read_all_doc(struct vc_data
cursor_track = read_all_mode;
spk_reset_index_count(0);
if (get_sentence_buf(vc, 0) == -1) {
- kbd_fakekey2(vc, RA_DOWN_ARROW);
+ del_timer(&cursor_timer);
+ if (!in_keyboard_notifier)
+ speakup_fake_down_arrow();
+ start_read_all_timer(vc, RA_DOWN_ARROW);
} else {
say_sentence_num(0, 0);
synth_insert_next_index(0);
@@ -2212,8 +2217,10 @@ static int keyboard_notifier_call(struct
int ret = NOTIFY_OK;
static int keycode; /* to hold the current keycode */
+ in_keyboard_notifier = 1;
+
if (vc->vc_mode == KD_GRAPHICS)
- return ret;
+ goto out;
/*
* First, determine whether we are handling a fake keypress on
@@ -2225,7 +2232,7 @@ static int keyboard_notifier_call(struct
*/
if (speakup_fake_key_pressed())
- return ret;
+ goto out;
switch (code) {
@@ -2266,6 +2273,8 @@ static int keyboard_notifier_call(struct
break;
}
}
+ in_keyboard_notifier = 0;
return ret;
}

--
How do
you spend it?
John Covici