hw/audio/sb16.c: missing break statement
PATCH 1/1 hw/audio/sb16.c: missing break statement
This patch adds a break in the switch() statement of complete(), value 0x42:
case 0x42: /* FT2 sets output freq with this, go figure */
qemu_log_mask(LOG_UNIMP, "cmd 0x42 might not do what it think it"
" should\n");
break; <-------
case 0x41:The issue was found by Coverity (#1385841):
CID 1385841: Control flow issues (MISSING_BREAK)
The case for value "66" is not terminated by a 'break' statement.
--- hw/audio/sb16.c | 1 + 1 file changed, 1 insertion(+)
hw/audio/sb16.c +1 -0
@@ -744,6 +744,7 @@ static void complete (SB16State *s)case 0x42 :/* FT2 sets output freq with this, go figure */ qemu_log_mask (LOG_UNIMP ,"cmd 0x42 might not do what it think it" " should\n" ); +break ;case 0x41 : s->freq =dsp_get_hilo (s);ldebug ("set freq %d\n" , s->freq);
Signature · 1 line
2.14.3
Hi Daniel,
Quoted 8 lines · Daniel Henrique Barboza
This patch adds a break in the
switch()statement ofcomplete(), value0x42:case 0x42: /* FT2 sets output freq with this, go figure */ qemu_log_mask(LOG_UNIMP, "cmd 0x42 might not do what it think it" " should\n"); break; <------- case 0x41:
It seems this is an intentional fallthrough, I understand cmd 0x42 is expected to do the same of 0x41 and _a bit more_ (see commit 85571bc7415).
Quoted 27 lines
The issue was found by Coverity (#1385841):
CID 1385841: Control flow issues (
MISSING_BREAK)
The case for value "66" is not terminated by a 'break' statement.Fixes 8ec660b80e ("hw/audio/sb16.c: change dolog() to qemu_log_mask()")Signed-off-by Daniel Henrique Barboza <[email protected]>CC John Arbuckle <[email protected]>CC Gerd Hoffmann <[email protected]>--- hw/audio/sb16.c | 1 + 1 file changed, 1 insertion(+)hw/audio/sb16.c +1
-0@@ -744,6 +744,7 @@ static void complete (SB16State *s)case 0x42 :/* FT2 sets output freq with this, go figure */ qemu_log_mask (LOG_UNIMP ,"cmd 0x42 might not do what it think it" " should\n" ); +break ;case 0x41 : s->freq =dsp_get_hilo (s);ldebug ("set freq %d\n" , s->freq);
Quoted 6 lines · Philippe Mathieu-Daudé
Hi Daniel,
Quoted 3 lines · Daniel Henrique Barboza
This patch adds a break in the
switch()statement ofcomplete(), value0x42:
Quoted 8 lines
Quoted 5 lines
case 0x42: /* FT2 sets output freq with this, go figure */ qemu_log_mask(LOG_UNIMP, "cmd 0x42 might not do what it think it" " should\n"); break; <------- case 0x41:It seems this is an intentional fallthrough, I understand cmd
0x42is expected to do the same of0x41and _a bitmore_(see commit85571bc7415).
Yes, I agree; I wrote a bit about this in this thread:
https://lists.gnu.org/archive/html/qemu-devel/2018-02/msg02081.html
(though my guess is that actually 0x42 is supposed to do exactly what 0x41 does, and that the LOG_UNIMP should maybe just be removed).
thanks
-- PMM
Quoted 13 lines · Peter Maydell
Quoted 12 lines · Philippe Mathieu-Daudé
Hi Daniel,
Quoted 8 lines · Daniel Henrique Barboza
This patch adds a break in the
switch()statement ofcomplete(), value0x42:case 0x42: /* FT2 sets output freq with this, go figure */ qemu_log_mask(LOG_UNIMP, "cmd 0x42 might not do what it think it" " should\n"); break; <------- case 0x41:
Quoted 5 lines
Quoted 2 lines
It seems this is an intentional fallthrough, I understand cmd
0x42is expected to do the same of0x41and _a bitmore_(see commit85571bc7415).Yes, I agree; I wrote a bit about this in this thread:
https://lists.gnu.org/archive/html/qemu-devel/2018-02/msg02081.html
Oh, very useful link!
Quoted 2 lines
(though my guess is that actually
0x42is supposed to do exactly what0x41does, and that theLOG_UNIMPshould maybe just be removed).
I now understand 0x42 sets the dsp input sampling freq, the model seems to be designed with output in mind, then added input support (using same freq as output).
So imho the simpler/safer fix would be:
case 0x42:
if (dsp_get_hilo(s) != s->freq) {
qemu_log_mask(LOG_UNIMP,
"input sampling freq different than "
"output not implemented");
}
/* fallthrough */
case 0x41:
...and the correct fix would be split s->freq in {s->freq_in, s->freq_out} but nobody ever required this during at least 14 years.
Quoted 20 lines · Philippe Mathieu-Daudé
Quoted 13 lines · Peter Maydell
Quoted 12 lines · Philippe Mathieu-Daudé
Hi Daniel,
Quoted 8 lines · Daniel Henrique Barboza
This patch adds a break in the
switch()statement ofcomplete(), value0x42:case 0x42: /* FT2 sets output freq with this, go figure */ qemu_log_mask(LOG_UNIMP, "cmd 0x42 might not do what it think it" " should\n"); break; <------- case 0x41:Quoted 5 lines
Quoted 2 lines
It seems this is an intentional fallthrough, I understand cmd
0x42is expected to do the same of0x41and _a bitmore_(see commit85571bc7415).Yes, I agree; I wrote a bit about this in this thread:
https://lists.gnu.org/archive/html/qemu-devel/2018-02/msg02081.html
Quoted 8 lines
Oh, very useful link!
Quoted 2 lines
(though my guess is that actually
0x42is supposed to do exactly what0x41does, and that theLOG_UNIMPshould maybe just be removed).I now understand
0x42sets the dsp input sampling freq, the model seems to be designed with output in mind, then added input support (using same freq as output).
Now I see Fabrice comment "FT2 sets output freq with this, go figure" and agree with him.
I like to think this is a bug in Fast Tracker 2, so Peter suggestion about using LOG_GUEST_ERROR here might be clever.
Quoted 16 lines
So imho the simpler/safer fix would be:
case 0x42: if (dsp_get_hilo(s) != s->freq) { qemu_log_mask(LOG_UNIMP, "input sampling freq different than " "output not implemented"); } /* fallthrough */ case 0x41: ...and the correct fix would be split
s->freqin {s->freq_in,s->freq_out} but nobody ever required this during at least 14 years.
Quoted 18 lines · Philippe Mathieu-Daudé
Now I see Fabrice comment "FT2 sets output freq with this, go figure" and agree with him.
I like to think this is a bug in Fast Tracker 2, so Peter suggestion about using
LOG_GUEST_ERRORhere might be clever.Quoted 12 lines
So imho the simpler/safer fix would be:
case 0x42: if (dsp_get_hilo(s) != s->freq) { qemu_log_mask(LOG_UNIMP, "input sampling freq different than " "output not implemented"); } /* fallthrough */ case 0x41: ...
Wouldn't this falsely report a warning for guest code that really is trying to set the input sampling frequency and doesn't care about output?
Quoted 1 line
Quoted 1 line
and the correct fix would be split
s->freqin {s->freq_in,s->freq_out}
...but that would differ from the hardware implementation, which (apparently) uses a single frequency for both.
thanks
-- PMM
Quoted 6 lines · Philippe Mathieu-Daudé
Hi Daniel,
Quoted 3 lines · Daniel Henrique Barboza
This patch adds a break in the
switch()statement ofcomplete(), value0x42:
Quoted 8 lines
Quoted 5 lines
case 0x42: /* FT2 sets output freq with this, go figure */ qemu_log_mask(LOG_UNIMP, "cmd 0x42 might not do what it think it" " should\n"); break; <------- case 0x41:It seems this is an intentional fallthrough, I understand cmd
0x42is expected to do the same of0x41and _a bitmore_(see commit85571bc7415).
It might be nice to turn on -Wimplicit-fallthrough and then annotate valid locations like this in qemu with /* fallthrough */
Although GCC has an __attribute((fallthrough)), the warning flag impl also looks for that magic comment, and the magic comment is portable to clang too.
Regards,
Daniel
Signature · 3 lines
|: https://berrange.com -o- https://www.flickr.com/photos/dberrange :| |: https://libvirt.org -o- https://fstop138.berrange.com :| |: https://entangle-photo.org -o- https://www.instagram.com/dberrange :|
Hey,
Quoted 6 lines · Philippe Mathieu-Daudé
Hi Daniel,
Quoted 3 lines · Daniel Henrique Barboza
This patch adds a break in the
switch()statement ofcomplete(), value0x42:
Quoted 7 lines
Quoted 5 lines
case 0x42: /* FT2 sets output freq with this, go figure */ qemu_log_mask(LOG_UNIMP, "cmd 0x42 might not do what it think it" " should\n"); break; <------- case 0x41:It seems this is an intentional fallthrough, I understand cmd
0x42is expected to do the same of0x41and _a bitmore_(see commit85571bc7415).
Perhaps it should be more explicit then? Something like
case 0x40:
s->time_const = dsp_get_data (s); ldebug ("set time const %d\n", s->time_const); break;
case 0x41:
case 0x42:
if (s->cmd == 0x42) {
/* FT2 sets output freq with this, go figure */ qemu_log_mask(LOG_UNIMP, "cmd 0x42 might not do what it think it" " should\n");
}
s->freq = dsp_get_hilo (s);
ldebug ("set freq %d\n", s->freq);
break;
case 0x48:
Quoted 27 lines
Quoted 27 lines
The issue was found by Coverity (#1385841):
CID 1385841: Control flow issues (
MISSING_BREAK)
The case for value "66" is not terminated by a 'break' statement.Fixes 8ec660b80e ("hw/audio/sb16.c: change dolog() to qemu_log_mask()")Signed-off-by Daniel Henrique Barboza <[email protected]>CC John Arbuckle <[email protected]>CC Gerd Hoffmann <[email protected]>--- hw/audio/sb16.c | 1 + 1 file changed, 1 insertion(+)hw/audio/sb16.c +1
-0@@ -744,6 +744,7 @@ static void complete (SB16State *s)case 0x42 :/* FT2 sets output freq with this, go figure */ qemu_log_mask (LOG_UNIMP ,"cmd 0x42 might not do what it think it" " should\n" ); +break ;case 0x41 : s->freq =dsp_get_hilo (s);ldebug ("set freq %d\n" , s->freq);