Skip to contentSkip to conversation

Optiarc DVD RW AD-5200A audio playing

DISCUSSION Applied Borislav Petkov 3 participants 12 messages Feb 16, 2008 – Feb 19, 2008 1 message not in the archive libata

lore
Share & tools
lore

A message that is not in the archive (sent off-list or not yet received): [email protected]

#1 lore raw
Quoted 14 lines · Stefan Bader

Hello Borislav,

I worked on a problem with an DVD driver (model=Optiarc DVD RW AD-5200A) which obviously has the same problem as some Matshita drives. The following patch was reported to enabled audio playing on this drive. Would this approach be suitable for upstream or are there other solutions to this problem?

Regards,
Stefan

drivers/ide/ide-cd.c int ide_cdrom_probe_capabilities (ide_drive_t *drive) · line 2995

Quoted 9 lines
drivers/ide/ide-cd.c +2 -1
@@ -2988,7 +2988,8 @@ int ide_cdrom_probe_capabilities (ide_drive_t *drive)
 	if (strcmp(drive->id->model, "MATSHITADVD-ROM SR-8187") == 0 ||
 	    strcmp(drive->id->model, "MATSHITADVD-ROM SR-8186") == 0 ||
 	    strcmp(drive->id->model, "MATSHITADVD-ROM SR-8176") == 0 ||
-	    strcmp(drive->id->model, "MATSHITADVD-ROM SR-8174") == 0)
+	    strcmp(drive->id->model, "MATSHITADVD-ROM SR-8174") == 0 ||
+	    strcmp(drive->id->model, "Optiarc DVD RW AD-5200A") == 0)
 		CDROM_CONFIG_FLAGS(drive)->audio_play = 1;
 
 #if ! STANDARD_ATAPI

Hi Stefan,

just to make sure that the audioplay bit is not set in the capabilities page, can you please try the following patch applied against 2.6.25-rc2 and send me the output. Thanks!

@Bart: by the way, this cdi->mask thingy is kinda unintuitive doing double negation to check whether a feature is supported or not. Yeah, this comes from "above," i.e. uniform cdrom layer. But still, shouldn't we use a cdi->caps_flags or something similar instead, which mirrors the caps page bits setting?

commit 435f0f4496a1b32af2d542f43b2370a890fe2f83

Author: Borislav Petkov <[email protected]>
Date:   Sat Feb 16 09:56:36 2008 +0100

ide-cd: Enable audio play quirk for Optiarc DVD RW AD-5200A drive

drivers/ide/ide-cd.c +5 -0
index f77db6b..2c9d06e 100644
@@ -1750,6 +1750,10 @@ int ide_cdrom_probe_capabilities (ide_drive_t *drive)
 		cdi->mask &= ~(CDC_DVD_RAM | CDC_RAM);
 	if (buf[8 + 3] & 0x10)
 		cdi->mask &= ~CDC_DVD_R;
+	if (!(buf[8 + 4] & 0x01)) {
+		printk(KERN_INFO "ide-cd: audio play not advertised in caps page,"
+				 " enabling quirk.\n");
+	}
 	if ((buf[8 + 4] & 0x01) || (cd->cd_flags & IDE_CD_FLAG_PLAY_AUDIO_OK))
 		cdi->mask &= ~CDC_PLAY_AUDIO;
 
@@ -1929,6 +1933,7 @@ static const struct cd_list_entry ide_cd_quirks_list[] = {
 	{ "MATSHITADVD-ROM SR-8186", NULL,   IDE_CD_FLAG_PLAY_AUDIO_OK	    },
 	{ "MATSHITADVD-ROM SR-8176", NULL,   IDE_CD_FLAG_PLAY_AUDIO_OK	    },
 	{ "MATSHITADVD-ROM SR-8174", NULL,   IDE_CD_FLAG_PLAY_AUDIO_OK	    },
+	{ "Optiarc DVD RW AD-5200A", NULL,   IDE_CD_FLAG_PLAY_AUDIO_OK      },
 	{ NULL, NULL, 0 }
 };
 
Signature · 2 lines
Regards/Gruß,
    Boris.
#2 lore raw

Hi,

Quoted 27 lines · Borislav Petkov
Quoted 14 lines · Stefan Bader

Hello Borislav,

I worked on a problem with an DVD driver (model=Optiarc DVD RW AD-5200A) which obviously has the same problem as some Matshita drives. The following patch was reported to enabled audio playing on this drive. Would this approach be suitable for upstream or are there other solutions to this problem?

Regards,
Stefan

drivers/ide/ide-cd.c int ide_cdrom_probe_capabilities (ide_drive_t *drive) · line 2995

Quoted 9 lines
drivers/ide/ide-cd.c +2 -1
@@ -2988,7 +2988,8 @@ int ide_cdrom_probe_capabilities (ide_drive_t *drive)
 	if (strcmp(drive->id->model, "MATSHITADVD-ROM SR-8187") == 0 ||
 	    strcmp(drive->id->model, "MATSHITADVD-ROM SR-8186") == 0 ||
 	    strcmp(drive->id->model, "MATSHITADVD-ROM SR-8176") == 0 ||
-	    strcmp(drive->id->model, "MATSHITADVD-ROM SR-8174") == 0)
+	    strcmp(drive->id->model, "MATSHITADVD-ROM SR-8174") == 0 ||
+	    strcmp(drive->id->model, "Optiarc DVD RW AD-5200A") == 0)
 		CDROM_CONFIG_FLAGS(drive)->audio_play = 1;
 
 #if ! STANDARD_ATAPI

Hi Stefan,

Quoted 8 lines

just to make sure that the audioplay bit is not set in the capabilities page, can you please try the following patch applied against 2.6.25-rc2 and send me the output. Thanks!

@Bart: by the way, this cdi->mask thingy is kinda unintuitive doing double negation to check whether a feature is supported or not. Yeah, this comes from "above," i.e. uniform cdrom layer. But still, shouldn't we use a cdi->caps_flags or something similar instead, which mirrors the caps page bits setting?

It seems so (at least having negative flags is very unintuitive) but they might be some reason for this ugliness, Jens?

[ Please also remember that since cdrom layer is _uniform_ it may be not
possible and/or desirable to have 1-1 mapping between caps page bits and the future cdi->caps_flags. ]

Quoted 14 lines

commit 435f0f4496a1b32af2d542f43b2370a890fe2f83

Author: Borislav Petkov <[email protected]>
Date:   Sat Feb 16 09:56:36 2008 +0100

ide-cd: Enable audio play quirk for Optiarc DVD RW AD-5200A drive

drivers/ide/ide-cd.c +0 -0
index f77db6b..2c9d06e 100644

drivers/ide/ide-cd.c int ide_cdrom_probe_capabilities (ide_drive_t *drive) · line 1753

Quoted 4 lines
drivers/ide/ide-cd.c +1 -0
@@ -1750,6 +1750,10 @@ int ide_cdrom_probe_capabilities (ide_drive_t *drive)
 		cdi->mask &= ~(CDC_DVD_RAM | CDC_RAM);
 	if (buf[8 + 3] & 0x10)
 		cdi->mask &= ~CDC_DVD_R;
+	if (!(buf[8 + 4] & 0x01)) {

Hmm, shouldn't there be '&& (cd->cd_flags & IDE_CD_FLAG_PLAY_AUDIO_OK)' to prevent false positives?

drivers/ide/ide-cd.c int ide_cdrom_probe_capabilities (ide_drive_t *drive) · line 1754

Quoted 1 line
drivers/ide/ide-cd.c +1 -0
+		printk(KERN_INFO "ide-cd: audio play not advertised in caps page,"

Would be nice to also printk() the device name.

Quoted 12 lines
drivers/ide/ide-cd.c +3 -0
+				 " enabling quirk.\n");
+	}
 	if ((buf[8 + 4] & 0x01) || (cd->cd_flags & IDE_CD_FLAG_PLAY_AUDIO_OK))
 		cdi->mask &= ~CDC_PLAY_AUDIO;
 
@@ -1929,6 +1933,7 @@ static const struct cd_list_entry ide_cd_quirks_list[] = {
 	{ "MATSHITADVD-ROM SR-8186", NULL,   IDE_CD_FLAG_PLAY_AUDIO_OK	    },
 	{ "MATSHITADVD-ROM SR-8176", NULL,   IDE_CD_FLAG_PLAY_AUDIO_OK	    },
 	{ "MATSHITADVD-ROM SR-8174", NULL,   IDE_CD_FLAG_PLAY_AUDIO_OK	    },
+	{ "Optiarc DVD RW AD-5200A", NULL,   IDE_CD_FLAG_PLAY_AUDIO_OK      },
 	{ NULL, NULL, 0 }
 };
#3 lore raw
Quoted 56 lines · Bartlomiej Zolnierkiewicz

Hi,

Quoted 27 lines · Borislav Petkov
Quoted 14 lines · Stefan Bader

Hello Borislav,

I worked on a problem with an DVD driver (model=Optiarc DVD RW AD-5200A) which obviously has the same problem as some Matshita drives. The following patch was reported to enabled audio playing on this drive. Would this approach be suitable for upstream or are there other solutions to this problem?

Regards,
Stefan

drivers/ide/ide-cd.c int ide_cdrom_probe_capabilities (ide_drive_t *drive) · line 2995

Quoted 9 lines
drivers/ide/ide-cd.c +2 -1
@@ -2988,7 +2988,8 @@ int ide_cdrom_probe_capabilities (ide_drive_t *drive)
 	if (strcmp(drive->id->model, "MATSHITADVD-ROM SR-8187") == 0 ||
 	    strcmp(drive->id->model, "MATSHITADVD-ROM SR-8186") == 0 ||
 	    strcmp(drive->id->model, "MATSHITADVD-ROM SR-8176") == 0 ||
-	    strcmp(drive->id->model, "MATSHITADVD-ROM SR-8174") == 0)
+	    strcmp(drive->id->model, "MATSHITADVD-ROM SR-8174") == 0 ||
+	    strcmp(drive->id->model, "Optiarc DVD RW AD-5200A") == 0)
 		CDROM_CONFIG_FLAGS(drive)->audio_play = 1;
 
 #if ! STANDARD_ATAPI

Hi Stefan,

Quoted 8 lines

just to make sure that the audioplay bit is not set in the capabilities page, can you please try the following patch applied against 2.6.25-rc2 and send me the output. Thanks!

@Bart: by the way, this cdi->mask thingy is kinda unintuitive doing double negation to check whether a feature is supported or not. Yeah, this comes from "above," i.e. uniform cdrom layer. But still, shouldn't we use a cdi->caps_flags or something similar instead, which mirrors the caps page bits setting?

It seems so (at least having negative flags is very unintuitive) but they might be some reason for this ugliness, Jens?

[ Please also remember that since cdrom layer is _uniform_ it may be not
possible and/or desirable to have 1-1 mapping between caps page bits and the future cdi->caps_flags. ]

Quoted 9 lines

commit 435f0f4496a1b32af2d542f43b2370a890fe2f83

Author: Borislav Petkov <[email protected]>
Date:   Sat Feb 16 09:56:36 2008 +0100

ide-cd: Enable audio play quirk for Optiarc DVD RW AD-5200A drive

Quoted 12 lines

drivers/ide/ide-cd.c int ide_cdrom_probe_capabilities (ide_drive_t *drive) · line 1753

Quoted 9 lines
drivers/ide/ide-cd.c +1 -0
index f77db6b..2c9d06e 100644
@@ -1750,6 +1750,10 @@ int ide_cdrom_probe_capabilities (ide_drive_t *drive)
 		cdi->mask &= ~(CDC_DVD_RAM | CDC_RAM);
 	if (buf[8 + 3] & 0x10)
 		cdi->mask &= ~CDC_DVD_R;
+	if (!(buf[8 + 4] & 0x01)) {

Hmm, shouldn't there be '&& (cd->cd_flags & IDE_CD_FLAG_PLAY_AUDIO_OK)' to prevent false positives?

I wanted to see whether the caps page reports the audioplay bit off...

Quoted 4 lines

drivers/ide/ide-cd.c int ide_cdrom_probe_capabilities (ide_drive_t *drive) · line 1754

Quoted 2 lines
drivers/ide/ide-cd.c +1 -0
+		printk(KERN_INFO "ide-cd: audio play not advertised in caps page,"

Would be nice to also printk() the device name.

... but printing the device model is actually a good idea and this will rule out false positives, so Stefan, please drop the previous patch and test the updated one below. Thanks.

commit 6cc44b0ce5c9270b15d456eb9ffa91b855e4e0d0

Author: Borislav Petkov <[email protected]>
Date:   Sat Feb 16 09:56:36 2008 +0100

ide-cd: Enable audio play quirk for Optiarc DVD RW AD-5200A drive

drivers/ide/ide-cd.c +6 -0
index f77db6b..4c9984f 100644
@@ -1750,6 +1750,11 @@ int ide_cdrom_probe_capabilities (ide_drive_t *drive)
 		cdi->mask &= ~(CDC_DVD_RAM | CDC_RAM);
 	if (buf[8 + 3] & 0x10)
 		cdi->mask &= ~CDC_DVD_R;
+	if (!(buf[8 + 4] & 0x01)) {
+		printk(KERN_INFO "ide-cd: audio play not advertised in caps "
+				"page for drive model [%s], enabling quirk.\n",
+				drive->id->model);
+	}
 	if ((buf[8 + 4] & 0x01) || (cd->cd_flags & IDE_CD_FLAG_PLAY_AUDIO_OK))
 		cdi->mask &= ~CDC_PLAY_AUDIO;
 
@@ -1929,6 +1934,7 @@ static const struct cd_list_entry ide_cd_quirks_list[] = {
 	{ "MATSHITADVD-ROM SR-8186", NULL,   IDE_CD_FLAG_PLAY_AUDIO_OK	    },
 	{ "MATSHITADVD-ROM SR-8176", NULL,   IDE_CD_FLAG_PLAY_AUDIO_OK	    },
 	{ "MATSHITADVD-ROM SR-8174", NULL,   IDE_CD_FLAG_PLAY_AUDIO_OK	    },
+	{ "Optiarc DVD RW AD-5200A", NULL,   IDE_CD_FLAG_PLAY_AUDIO_OK      },
 	{ NULL, NULL, 0 }
 };
 
Signature · 2 lines
Regards/Gruß,
    Boris.
#4 lore raw
Quoted 70 lines · Borislav Petkov
Quoted 56 lines · Bartlomiej Zolnierkiewicz

Hi,

Quoted 27 lines · Borislav Petkov
Quoted 14 lines · Stefan Bader

Hello Borislav,

I worked on a problem with an DVD driver (model=Optiarc DVD RW AD-5200A) which obviously has the same problem as some Matshita drives. The following patch was reported to enabled audio playing on this drive. Would this approach be suitable for upstream or are there other solutions to this problem?

Regards,
Stefan

drivers/ide/ide-cd.c int ide_cdrom_probe_capabilities (ide_drive_t *drive) · line 2995

Quoted 9 lines
drivers/ide/ide-cd.c +2 -1
@@ -2988,7 +2988,8 @@ int ide_cdrom_probe_capabilities (ide_drive_t *drive)
 	if (strcmp(drive->id->model, "MATSHITADVD-ROM SR-8187") == 0 ||
 	    strcmp(drive->id->model, "MATSHITADVD-ROM SR-8186") == 0 ||
 	    strcmp(drive->id->model, "MATSHITADVD-ROM SR-8176") == 0 ||
-	    strcmp(drive->id->model, "MATSHITADVD-ROM SR-8174") == 0)
+	    strcmp(drive->id->model, "MATSHITADVD-ROM SR-8174") == 0 ||
+	    strcmp(drive->id->model, "Optiarc DVD RW AD-5200A") == 0)
 		CDROM_CONFIG_FLAGS(drive)->audio_play = 1;
 
 #if ! STANDARD_ATAPI

Hi Stefan,

Quoted 8 lines

just to make sure that the audioplay bit is not set in the capabilities page, can you please try the following patch applied against 2.6.25-rc2 and send me the output. Thanks!

@Bart: by the way, this cdi->mask thingy is kinda unintuitive doing double negation to check whether a feature is supported or not. Yeah, this comes from "above," i.e. uniform cdrom layer. But still, shouldn't we use a cdi->caps_flags or something similar instead, which mirrors the caps page bits setting?

It seems so (at least having negative flags is very unintuitive) but they might be some reason for this ugliness, Jens?

[ Please also remember that since cdrom layer is _uniform_ it may be not
possible and/or desirable to have 1-1 mapping between caps page bits and the future cdi->caps_flags. ]

Quoted 9 lines

commit 435f0f4496a1b32af2d542f43b2370a890fe2f83

Author: Borislav Petkov <[email protected]>
Date:   Sat Feb 16 09:56:36 2008 +0100

ide-cd: Enable audio play quirk for Optiarc DVD RW AD-5200A drive

Quoted 12 lines

drivers/ide/ide-cd.c int ide_cdrom_probe_capabilities (ide_drive_t *drive) · line 1753

Quoted 9 lines
drivers/ide/ide-cd.c +1 -0
index f77db6b..2c9d06e 100644
@@ -1750,6 +1750,10 @@ int ide_cdrom_probe_capabilities (ide_drive_t *drive)
 		cdi->mask &= ~(CDC_DVD_RAM | CDC_RAM);
 	if (buf[8 + 3] & 0x10)
 		cdi->mask &= ~CDC_DVD_R;
+	if (!(buf[8 + 4] & 0x01)) {

Hmm, shouldn't there be '&& (cd->cd_flags & IDE_CD_FLAG_PLAY_AUDIO_OK)' to prevent false positives?

Quoted 1 line

I wanted to see whether the caps page reports the audioplay bit off...

we still need IDE_CD_FLAG_PLAY_AUDIO_OK flag to be _set_ to enable the quirk

Quoted 24 lines
Quoted 4 lines

drivers/ide/ide-cd.c int ide_cdrom_probe_capabilities (ide_drive_t *drive) · line 1754

Quoted 2 lines
drivers/ide/ide-cd.c +1 -0
+		printk(KERN_INFO "ide-cd: audio play not advertised in caps page,"

Would be nice to also printk() the device name.

... but printing the device model is actually a good idea and this will rule out false positives, so Stefan, please drop the previous patch and test the updated one below. Thanks.

commit 6cc44b0ce5c9270b15d456eb9ffa91b855e4e0d0

Author: Borislav Petkov <[email protected]>
Date:   Sat Feb 16 09:56:36 2008 +0100

ide-cd: Enable audio play quirk for Optiarc DVD RW AD-5200A drive

drivers/ide/ide-cd.c +0 -0
index f77db6b..4c9984f 100644

drivers/ide/ide-cd.c int ide_cdrom_probe_capabilities (ide_drive_t *drive) · line 1756

Quoted 7 lines
drivers/ide/ide-cd.c +4 -0
@@ -1750,6 +1750,11 @@ int ide_cdrom_probe_capabilities (ide_drive_t *drive)
 		cdi->mask &= ~(CDC_DVD_RAM | CDC_RAM);
 	if (buf[8 + 3] & 0x10)
 		cdi->mask &= ~CDC_DVD_R;
+	if (!(buf[8 + 4] & 0x01)) {
+		printk(KERN_INFO "ide-cd: audio play not advertised in caps "
+				"page for drive model [%s], enabling quirk.\n",
+				drive->id->model);

if IDE_CD_FLAG_PLAY_AUDIO_OK flag is not set the above message is _bogus_ (because the quirk won't be enabled)

[ how's about just deleting the whole printk() to preserve simplicity? ]

Quoted 11 lines
drivers/ide/ide-cd.c +2 -0
+	}
 	if ((buf[8 + 4] & 0x01) || (cd->cd_flags & IDE_CD_FLAG_PLAY_AUDIO_OK))
 		cdi->mask &= ~CDC_PLAY_AUDIO;
 
@@ -1929,6 +1934,7 @@ static const struct cd_list_entry ide_cd_quirks_list[] = {
 	{ "MATSHITADVD-ROM SR-8186", NULL,   IDE_CD_FLAG_PLAY_AUDIO_OK	    },
 	{ "MATSHITADVD-ROM SR-8176", NULL,   IDE_CD_FLAG_PLAY_AUDIO_OK	    },
 	{ "MATSHITADVD-ROM SR-8174", NULL,   IDE_CD_FLAG_PLAY_AUDIO_OK	    },
+	{ "Optiarc DVD RW AD-5200A", NULL,   IDE_CD_FLAG_PLAY_AUDIO_OK      },
 	{ NULL, NULL, 0 }
 };
#5 lore raw
Quoted 116 lines · Bartlomiej Zolnierkiewicz
Quoted 70 lines · Borislav Petkov
Quoted 56 lines · Bartlomiej Zolnierkiewicz

Hi,

Quoted 27 lines · Borislav Petkov
Quoted 14 lines · Stefan Bader

Hello Borislav,

I worked on a problem with an DVD driver (model=Optiarc DVD RW AD-5200A) which obviously has the same problem as some Matshita drives. The following patch was reported to enabled audio playing on this drive. Would this approach be suitable for upstream or are there other solutions to this problem?

Regards,
Stefan

drivers/ide/ide-cd.c int ide_cdrom_probe_capabilities (ide_drive_t *drive) · line 2995

Quoted 9 lines
drivers/ide/ide-cd.c +2 -1
@@ -2988,7 +2988,8 @@ int ide_cdrom_probe_capabilities (ide_drive_t *drive)
 	if (strcmp(drive->id->model, "MATSHITADVD-ROM SR-8187") == 0 ||
 	    strcmp(drive->id->model, "MATSHITADVD-ROM SR-8186") == 0 ||
 	    strcmp(drive->id->model, "MATSHITADVD-ROM SR-8176") == 0 ||
-	    strcmp(drive->id->model, "MATSHITADVD-ROM SR-8174") == 0)
+	    strcmp(drive->id->model, "MATSHITADVD-ROM SR-8174") == 0 ||
+	    strcmp(drive->id->model, "Optiarc DVD RW AD-5200A") == 0)
 		CDROM_CONFIG_FLAGS(drive)->audio_play = 1;
 
 #if ! STANDARD_ATAPI

Hi Stefan,

Quoted 8 lines

just to make sure that the audioplay bit is not set in the capabilities page, can you please try the following patch applied against 2.6.25-rc2 and send me the output. Thanks!

@Bart: by the way, this cdi->mask thingy is kinda unintuitive doing double negation to check whether a feature is supported or not. Yeah, this comes from "above," i.e. uniform cdrom layer. But still, shouldn't we use a cdi->caps_flags or something similar instead, which mirrors the caps page bits setting?

It seems so (at least having negative flags is very unintuitive) but they might be some reason for this ugliness, Jens?

[ Please also remember that since cdrom layer is _uniform_ it may be not
possible and/or desirable to have 1-1 mapping between caps page bits and the future cdi->caps_flags. ]

Quoted 9 lines

commit 435f0f4496a1b32af2d542f43b2370a890fe2f83

Author: Borislav Petkov <[email protected]>
Date:   Sat Feb 16 09:56:36 2008 +0100

ide-cd: Enable audio play quirk for Optiarc DVD RW AD-5200A drive

Quoted 12 lines

drivers/ide/ide-cd.c int ide_cdrom_probe_capabilities (ide_drive_t *drive) · line 1753

Quoted 9 lines
drivers/ide/ide-cd.c +1 -0
index f77db6b..2c9d06e 100644
@@ -1750,6 +1750,10 @@ int ide_cdrom_probe_capabilities (ide_drive_t *drive)
 		cdi->mask &= ~(CDC_DVD_RAM | CDC_RAM);
 	if (buf[8 + 3] & 0x10)
 		cdi->mask &= ~CDC_DVD_R;
+	if (!(buf[8 + 4] & 0x01)) {

Hmm, shouldn't there be '&& (cd->cd_flags & IDE_CD_FLAG_PLAY_AUDIO_OK)' to prevent false positives?

Quoted 1 line

I wanted to see whether the caps page reports the audioplay bit off...

we still need IDE_CD_FLAG_PLAY_AUDIO_OK flag to be _set_ to enable the quirk

Quoted 24 lines
Quoted 4 lines

drivers/ide/ide-cd.c int ide_cdrom_probe_capabilities (ide_drive_t *drive) · line 1754

Quoted 2 lines
drivers/ide/ide-cd.c +1 -0
+		printk(KERN_INFO "ide-cd: audio play not advertised in caps page,"

Would be nice to also printk() the device name.

... but printing the device model is actually a good idea and this will rule out false positives, so Stefan, please drop the previous patch and test the updated one below. Thanks.

commit 6cc44b0ce5c9270b15d456eb9ffa91b855e4e0d0

Author: Borislav Petkov <[email protected]>
Date:   Sat Feb 16 09:56:36 2008 +0100

ide-cd: Enable audio play quirk for Optiarc DVD RW AD-5200A drive

drivers/ide/ide-cd.c +0 -0
index f77db6b..4c9984f 100644

drivers/ide/ide-cd.c int ide_cdrom_probe_capabilities (ide_drive_t *drive) · line 1756

Quoted 7 lines
drivers/ide/ide-cd.c +4 -0
@@ -1750,6 +1750,11 @@ int ide_cdrom_probe_capabilities (ide_drive_t *drive)
 		cdi->mask &= ~(CDC_DVD_RAM | CDC_RAM);
 	if (buf[8 + 3] & 0x10)
 		cdi->mask &= ~CDC_DVD_R;
+	if (!(buf[8 + 4] & 0x01)) {
+		printk(KERN_INFO "ide-cd: audio play not advertised in caps "
+				"page for drive model [%s], enabling quirk.\n",
+				drive->id->model);

if IDE_CD_FLAG_PLAY_AUDIO_OK flag is not set the above message is _bogus_ (because the quirk won't be enabled)

[ how's about just deleting the whole printk() to preserve simplicity? ]

Quoted 4 lines
drivers/ide/ide-cd.c +1 -0
+	}
 	if ((buf[8 + 4] & 0x01) || (cd->cd_flags & IDE_CD_FLAG_PLAY_AUDIO_OK))
 		cdi->mask &= ~CDC_PLAY_AUDIO;
 
Quoted 5 lines

drivers/ide/ide-cd.c static const struct cd_list_entry ide_cd_quirks_list[] = { · line 1937

Quoted 5 lines
drivers/ide/ide-cd.c +1 -0
@@ -1929,6 +1934,7 @@ static const struct cd_list_entry ide_cd_quirks_list[] = {
 	{ "MATSHITADVD-ROM SR-8186", NULL,   IDE_CD_FLAG_PLAY_AUDIO_OK	    },
 	{ "MATSHITADVD-ROM SR-8176", NULL,   IDE_CD_FLAG_PLAY_AUDIO_OK	    },
 	{ "MATSHITADVD-ROM SR-8174", NULL,   IDE_CD_FLAG_PLAY_AUDIO_OK	    },
+	{ "Optiarc DVD RW AD-5200A", NULL,   IDE_CD_FLAG_PLAY_AUDIO_OK      },

But the flag _is_ set because the drive is added to the quirk list (line above ^) and
ide_cdrom_setup() calls ide_cd_flags() to set the proper flags for cd->cd_flags and later on ide_cdrom_probe_capabilities() works on the same cd->cd_flags for testing IDE_CD_FLAG_PLAY_AUDIO_OK, or have i gone blind completely...?

Quoted 2 lines
Quoted 2 lines
drivers/ide/ide-cd.c +0 -0
 	{ NULL, NULL, 0 }
 };
Signature · 2 lines
Regards/Gruß,
    Boris.
#6 lore raw
Quoted 117 lines · Borislav Petkov
Quoted 116 lines · Bartlomiej Zolnierkiewicz
Quoted 70 lines · Borislav Petkov
Quoted 56 lines · Bartlomiej Zolnierkiewicz

Hi,

Quoted 27 lines · Borislav Petkov
Quoted 14 lines · Stefan Bader

Hello Borislav,

I worked on a problem with an DVD driver (model=Optiarc DVD RW AD-5200A) which obviously has the same problem as some Matshita drives. The following patch was reported to enabled audio playing on this drive. Would this approach be suitable for upstream or are there other solutions to this problem?

Regards,
Stefan

drivers/ide/ide-cd.c int ide_cdrom_probe_capabilities (ide_drive_t *drive) · line 2995

Quoted 9 lines
drivers/ide/ide-cd.c +2 -1
@@ -2988,7 +2988,8 @@ int ide_cdrom_probe_capabilities (ide_drive_t *drive)
 	if (strcmp(drive->id->model, "MATSHITADVD-ROM SR-8187") == 0 ||
 	    strcmp(drive->id->model, "MATSHITADVD-ROM SR-8186") == 0 ||
 	    strcmp(drive->id->model, "MATSHITADVD-ROM SR-8176") == 0 ||
-	    strcmp(drive->id->model, "MATSHITADVD-ROM SR-8174") == 0)
+	    strcmp(drive->id->model, "MATSHITADVD-ROM SR-8174") == 0 ||
+	    strcmp(drive->id->model, "Optiarc DVD RW AD-5200A") == 0)
 		CDROM_CONFIG_FLAGS(drive)->audio_play = 1;
 
 #if ! STANDARD_ATAPI

Hi Stefan,

Quoted 8 lines

just to make sure that the audioplay bit is not set in the capabilities page, can you please try the following patch applied against 2.6.25-rc2 and send me the output. Thanks!

@Bart: by the way, this cdi->mask thingy is kinda unintuitive doing double negation to check whether a feature is supported or not. Yeah, this comes from "above," i.e. uniform cdrom layer. But still, shouldn't we use a cdi->caps_flags or something similar instead, which mirrors the caps page bits setting?

It seems so (at least having negative flags is very unintuitive) but they might be some reason for this ugliness, Jens?

[ Please also remember that since cdrom layer is _uniform_ it may be not
possible and/or desirable to have 1-1 mapping between caps page bits and the future cdi->caps_flags. ]

Quoted 9 lines

commit 435f0f4496a1b32af2d542f43b2370a890fe2f83

Author: Borislav Petkov <[email protected]>
Date:   Sat Feb 16 09:56:36 2008 +0100

ide-cd: Enable audio play quirk for Optiarc DVD RW AD-5200A drive

Quoted 12 lines

drivers/ide/ide-cd.c int ide_cdrom_probe_capabilities (ide_drive_t *drive) · line 1753

Quoted 9 lines
drivers/ide/ide-cd.c +1 -0
index f77db6b..2c9d06e 100644
@@ -1750,6 +1750,10 @@ int ide_cdrom_probe_capabilities (ide_drive_t *drive)
 		cdi->mask &= ~(CDC_DVD_RAM | CDC_RAM);
 	if (buf[8 + 3] & 0x10)
 		cdi->mask &= ~CDC_DVD_R;
+	if (!(buf[8 + 4] & 0x01)) {

Hmm, shouldn't there be '&& (cd->cd_flags & IDE_CD_FLAG_PLAY_AUDIO_OK)' to prevent false positives?

Quoted 1 line

I wanted to see whether the caps page reports the audioplay bit off...

we still need IDE_CD_FLAG_PLAY_AUDIO_OK flag to be _set_ to enable the quirk

Quoted 24 lines
Quoted 4 lines

drivers/ide/ide-cd.c int ide_cdrom_probe_capabilities (ide_drive_t *drive) · line 1754

Quoted 2 lines
drivers/ide/ide-cd.c +1 -0
+		printk(KERN_INFO "ide-cd: audio play not advertised in caps page,"

Would be nice to also printk() the device name.

... but printing the device model is actually a good idea and this will rule out false positives, so Stefan, please drop the previous patch and test the updated one below. Thanks.

commit 6cc44b0ce5c9270b15d456eb9ffa91b855e4e0d0

Author: Borislav Petkov <[email protected]>
Date:   Sat Feb 16 09:56:36 2008 +0100

ide-cd: Enable audio play quirk for Optiarc DVD RW AD-5200A drive

drivers/ide/ide-cd.c +0 -0
index f77db6b..4c9984f 100644

drivers/ide/ide-cd.c int ide_cdrom_probe_capabilities (ide_drive_t *drive) · line 1756

Quoted 7 lines
drivers/ide/ide-cd.c +4 -0
@@ -1750,6 +1750,11 @@ int ide_cdrom_probe_capabilities (ide_drive_t *drive)
 		cdi->mask &= ~(CDC_DVD_RAM | CDC_RAM);
 	if (buf[8 + 3] & 0x10)
 		cdi->mask &= ~CDC_DVD_R;
+	if (!(buf[8 + 4] & 0x01)) {
+		printk(KERN_INFO "ide-cd: audio play not advertised in caps "
+				"page for drive model [%s], enabling quirk.\n",
+				drive->id->model);

if IDE_CD_FLAG_PLAY_AUDIO_OK flag is not set the above message is _bogus_ (because the quirk won't be enabled)

[ how's about just deleting the whole printk() to preserve simplicity? ]

Quoted 4 lines
drivers/ide/ide-cd.c +1 -0
+	}
 	if ((buf[8 + 4] & 0x01) || (cd->cd_flags & IDE_CD_FLAG_PLAY_AUDIO_OK))
 		cdi->mask &= ~CDC_PLAY_AUDIO;
 
Quoted 7 lines
Quoted 5 lines

drivers/ide/ide-cd.c static const struct cd_list_entry ide_cd_quirks_list[] = { · line 1937

Quoted 5 lines
drivers/ide/ide-cd.c +1 -0
@@ -1929,6 +1934,7 @@ static const struct cd_list_entry ide_cd_quirks_list[] = {
 	{ "MATSHITADVD-ROM SR-8186", NULL,   IDE_CD_FLAG_PLAY_AUDIO_OK	    },
 	{ "MATSHITADVD-ROM SR-8176", NULL,   IDE_CD_FLAG_PLAY_AUDIO_OK	    },
 	{ "MATSHITADVD-ROM SR-8174", NULL,   IDE_CD_FLAG_PLAY_AUDIO_OK	    },
+	{ "Optiarc DVD RW AD-5200A", NULL,   IDE_CD_FLAG_PLAY_AUDIO_OK      },

But the flag _is_ set because the drive is added to the quirk list (line above ^) and

                                ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

(fortunately) majority of ATAPI drives are not in the quirk list ;-)

[ and they may have (buf[8 + 4] & 0x01) == 0 ]

Quoted 6 lines

ide_cdrom_setup() calls ide_cd_flags() to set the proper flags for cd->cd_flags and later on ide_cdrom_probe_capabilities() works on the same cd->cd_flags for testing IDE_CD_FLAG_PLAY_AUDIO_OK, or have i gone blind completely...?

Quoted 2 lines
Quoted 2 lines
drivers/ide/ide-cd.c +0 -0
 	{ NULL, NULL, 0 }
 };
#7 lore raw
Quoted 125 lines · Bartlomiej Zolnierkiewicz
Quoted 117 lines · Borislav Petkov
Quoted 116 lines · Bartlomiej Zolnierkiewicz
Quoted 70 lines · Borislav Petkov
Quoted 56 lines · Bartlomiej Zolnierkiewicz

Hi,

Quoted 27 lines · Borislav Petkov
Quoted 14 lines · Stefan Bader

Hello Borislav,

I worked on a problem with an DVD driver (model=Optiarc DVD RW AD-5200A) which obviously has the same problem as some Matshita drives. The following patch was reported to enabled audio playing on this drive. Would this approach be suitable for upstream or are there other solutions to this problem?

Regards,
Stefan

drivers/ide/ide-cd.c int ide_cdrom_probe_capabilities (ide_drive_t *drive) · line 2995

Quoted 9 lines
drivers/ide/ide-cd.c +2 -1
@@ -2988,7 +2988,8 @@ int ide_cdrom_probe_capabilities (ide_drive_t *drive)
 	if (strcmp(drive->id->model, "MATSHITADVD-ROM SR-8187") == 0 ||
 	    strcmp(drive->id->model, "MATSHITADVD-ROM SR-8186") == 0 ||
 	    strcmp(drive->id->model, "MATSHITADVD-ROM SR-8176") == 0 ||
-	    strcmp(drive->id->model, "MATSHITADVD-ROM SR-8174") == 0)
+	    strcmp(drive->id->model, "MATSHITADVD-ROM SR-8174") == 0 ||
+	    strcmp(drive->id->model, "Optiarc DVD RW AD-5200A") == 0)
 		CDROM_CONFIG_FLAGS(drive)->audio_play = 1;
 
 #if ! STANDARD_ATAPI

Hi Stefan,

Quoted 8 lines

just to make sure that the audioplay bit is not set in the capabilities page, can you please try the following patch applied against 2.6.25-rc2 and send me the output. Thanks!

@Bart: by the way, this cdi->mask thingy is kinda unintuitive doing double negation to check whether a feature is supported or not. Yeah, this comes from "above," i.e. uniform cdrom layer. But still, shouldn't we use a cdi->caps_flags or something similar instead, which mirrors the caps page bits setting?

It seems so (at least having negative flags is very unintuitive) but they might be some reason for this ugliness, Jens?

[ Please also remember that since cdrom layer is _uniform_ it may be not
possible and/or desirable to have 1-1 mapping between caps page bits and the future cdi->caps_flags. ]

Quoted 9 lines

commit 435f0f4496a1b32af2d542f43b2370a890fe2f83

Author: Borislav Petkov <[email protected]>
Date:   Sat Feb 16 09:56:36 2008 +0100

ide-cd: Enable audio play quirk for Optiarc DVD RW AD-5200A drive

Quoted 12 lines

drivers/ide/ide-cd.c int ide_cdrom_probe_capabilities (ide_drive_t *drive) · line 1753

Quoted 9 lines
drivers/ide/ide-cd.c +1 -0
index f77db6b..2c9d06e 100644
@@ -1750,6 +1750,10 @@ int ide_cdrom_probe_capabilities (ide_drive_t *drive)
 		cdi->mask &= ~(CDC_DVD_RAM | CDC_RAM);
 	if (buf[8 + 3] & 0x10)
 		cdi->mask &= ~CDC_DVD_R;
+	if (!(buf[8 + 4] & 0x01)) {

Hmm, shouldn't there be '&& (cd->cd_flags & IDE_CD_FLAG_PLAY_AUDIO_OK)' to prevent false positives?

Quoted 1 line

I wanted to see whether the caps page reports the audioplay bit off...

we still need IDE_CD_FLAG_PLAY_AUDIO_OK flag to be _set_ to enable the quirk

Quoted 24 lines
Quoted 4 lines

drivers/ide/ide-cd.c int ide_cdrom_probe_capabilities (ide_drive_t *drive) · line 1754

Quoted 2 lines
drivers/ide/ide-cd.c +1 -0
+		printk(KERN_INFO "ide-cd: audio play not advertised in caps page,"

Would be nice to also printk() the device name.

... but printing the device model is actually a good idea and this will rule out false positives, so Stefan, please drop the previous patch and test the updated one below. Thanks.

commit 6cc44b0ce5c9270b15d456eb9ffa91b855e4e0d0

Author: Borislav Petkov <[email protected]>
Date:   Sat Feb 16 09:56:36 2008 +0100

ide-cd: Enable audio play quirk for Optiarc DVD RW AD-5200A drive

drivers/ide/ide-cd.c +0 -0
index f77db6b..4c9984f 100644

drivers/ide/ide-cd.c int ide_cdrom_probe_capabilities (ide_drive_t *drive) · line 1756

Quoted 7 lines
drivers/ide/ide-cd.c +4 -0
@@ -1750,6 +1750,11 @@ int ide_cdrom_probe_capabilities (ide_drive_t *drive)
 		cdi->mask &= ~(CDC_DVD_RAM | CDC_RAM);
 	if (buf[8 + 3] & 0x10)
 		cdi->mask &= ~CDC_DVD_R;
+	if (!(buf[8 + 4] & 0x01)) {
+		printk(KERN_INFO "ide-cd: audio play not advertised in caps "
+				"page for drive model [%s], enabling quirk.\n",
+				drive->id->model);

if IDE_CD_FLAG_PLAY_AUDIO_OK flag is not set the above message is _bogus_ (because the quirk won't be enabled)

[ how's about just deleting the whole printk() to preserve simplicity? ]

Quoted 4 lines
drivers/ide/ide-cd.c +1 -0
+	}
 	if ((buf[8 + 4] & 0x01) || (cd->cd_flags & IDE_CD_FLAG_PLAY_AUDIO_OK))
 		cdi->mask &= ~CDC_PLAY_AUDIO;
 
Quoted 7 lines
Quoted 5 lines

drivers/ide/ide-cd.c static const struct cd_list_entry ide_cd_quirks_list[] = { · line 1937

Quoted 5 lines
drivers/ide/ide-cd.c +1 -0
@@ -1929,6 +1934,7 @@ static const struct cd_list_entry ide_cd_quirks_list[] = {
 	{ "MATSHITADVD-ROM SR-8186", NULL,   IDE_CD_FLAG_PLAY_AUDIO_OK	    },
 	{ "MATSHITADVD-ROM SR-8176", NULL,   IDE_CD_FLAG_PLAY_AUDIO_OK	    },
 	{ "MATSHITADVD-ROM SR-8174", NULL,   IDE_CD_FLAG_PLAY_AUDIO_OK	    },
+	{ "Optiarc DVD RW AD-5200A", NULL,   IDE_CD_FLAG_PLAY_AUDIO_OK      },

But the flag _is_ set because the drive is added to the quirk list (line above ^) and

Quoted 5 lines
                                ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

(fortunately) majority of ATAPI drives are not in the quirk list ;-)

[ and they may have (buf[8 + 4] & 0x01) == 0 ]

Bart,

i think i probably didn't express myself clear enough before. The patch serves the sole purpose to show exactly whether buf[8 + 4] & 0x01) == 0 and is in no way the final solution. I just wanted to make sure the caps page doesn't set the audioplay bit. The final solution would be to add the drive to the quirks list only, i guess, in case you don't have a better idea.

Sorry for the misunderstanding :=).

Quoted 7 lines
Quoted 7 lines

ide_cdrom_setup() calls ide_cd_flags() to set the proper flags for cd->cd_flags and later on ide_cdrom_probe_capabilities() works on the same cd->cd_flags for testing IDE_CD_FLAG_PLAY_AUDIO_OK, or have i gone blind completely...?

Quoted 2 lines
Quoted 2 lines
drivers/ide/ide-cd.c +0 -0
 	{ NULL, NULL, 0 }
 };
Signature · 2 lines
Regards/Gruß,
    Boris.
#8 lore raw
Quoted 134 lines · Borislav Petkov
Quoted 125 lines · Bartlomiej Zolnierkiewicz
Quoted 117 lines · Borislav Petkov
Quoted 116 lines · Bartlomiej Zolnierkiewicz
Quoted 70 lines · Borislav Petkov
Quoted 56 lines · Bartlomiej Zolnierkiewicz

Hi,

Quoted 27 lines · Borislav Petkov
Quoted 14 lines · Stefan Bader

Hello Borislav,

I worked on a problem with an DVD driver (model=Optiarc DVD RW AD-5200A) which obviously has the same problem as some Matshita drives. The following patch was reported to enabled audio playing on this drive. Would this approach be suitable for upstream or are there other solutions to this problem?

Regards,
Stefan

drivers/ide/ide-cd.c int ide_cdrom_probe_capabilities (ide_drive_t *drive) · line 2995

Quoted 9 lines
drivers/ide/ide-cd.c +2 -1
@@ -2988,7 +2988,8 @@ int ide_cdrom_probe_capabilities (ide_drive_t *drive)
 	if (strcmp(drive->id->model, "MATSHITADVD-ROM SR-8187") == 0 ||
 	    strcmp(drive->id->model, "MATSHITADVD-ROM SR-8186") == 0 ||
 	    strcmp(drive->id->model, "MATSHITADVD-ROM SR-8176") == 0 ||
-	    strcmp(drive->id->model, "MATSHITADVD-ROM SR-8174") == 0)
+	    strcmp(drive->id->model, "MATSHITADVD-ROM SR-8174") == 0 ||
+	    strcmp(drive->id->model, "Optiarc DVD RW AD-5200A") == 0)
 		CDROM_CONFIG_FLAGS(drive)->audio_play = 1;
 
 #if ! STANDARD_ATAPI

Hi Stefan,

Quoted 8 lines

just to make sure that the audioplay bit is not set in the capabilities page, can you please try the following patch applied against 2.6.25-rc2 and send me the output. Thanks!

@Bart: by the way, this cdi->mask thingy is kinda unintuitive doing double negation to check whether a feature is supported or not. Yeah, this comes from "above," i.e. uniform cdrom layer. But still, shouldn't we use a cdi->caps_flags or something similar instead, which mirrors the caps page bits setting?

It seems so (at least having negative flags is very unintuitive) but they might be some reason for this ugliness, Jens?

[ Please also remember that since cdrom layer is _uniform_ it may be not
possible and/or desirable to have 1-1 mapping between caps page bits and the future cdi->caps_flags. ]

Quoted 9 lines

commit 435f0f4496a1b32af2d542f43b2370a890fe2f83

Author: Borislav Petkov <[email protected]>
Date:   Sat Feb 16 09:56:36 2008 +0100

ide-cd: Enable audio play quirk for Optiarc DVD RW AD-5200A drive

Quoted 12 lines

drivers/ide/ide-cd.c int ide_cdrom_probe_capabilities (ide_drive_t *drive) · line 1753

Quoted 9 lines
drivers/ide/ide-cd.c +1 -0
index f77db6b..2c9d06e 100644
@@ -1750,6 +1750,10 @@ int ide_cdrom_probe_capabilities (ide_drive_t *drive)
 		cdi->mask &= ~(CDC_DVD_RAM | CDC_RAM);
 	if (buf[8 + 3] & 0x10)
 		cdi->mask &= ~CDC_DVD_R;
+	if (!(buf[8 + 4] & 0x01)) {

Hmm, shouldn't there be '&& (cd->cd_flags & IDE_CD_FLAG_PLAY_AUDIO_OK)' to prevent false positives?

Quoted 1 line

I wanted to see whether the caps page reports the audioplay bit off...

we still need IDE_CD_FLAG_PLAY_AUDIO_OK flag to be _set_ to enable the quirk

Quoted 24 lines
Quoted 4 lines

drivers/ide/ide-cd.c int ide_cdrom_probe_capabilities (ide_drive_t *drive) · line 1754

Quoted 2 lines
drivers/ide/ide-cd.c +1 -0
+		printk(KERN_INFO "ide-cd: audio play not advertised in caps page,"

Would be nice to also printk() the device name.

... but printing the device model is actually a good idea and this will rule out false positives, so Stefan, please drop the previous patch and test the updated one below. Thanks.

commit 6cc44b0ce5c9270b15d456eb9ffa91b855e4e0d0

Author: Borislav Petkov <[email protected]>
Date:   Sat Feb 16 09:56:36 2008 +0100

ide-cd: Enable audio play quirk for Optiarc DVD RW AD-5200A drive

drivers/ide/ide-cd.c +0 -0
index f77db6b..4c9984f 100644

drivers/ide/ide-cd.c int ide_cdrom_probe_capabilities (ide_drive_t *drive) · line 1756

Quoted 7 lines
drivers/ide/ide-cd.c +4 -0
@@ -1750,6 +1750,11 @@ int ide_cdrom_probe_capabilities (ide_drive_t *drive)
 		cdi->mask &= ~(CDC_DVD_RAM | CDC_RAM);
 	if (buf[8 + 3] & 0x10)
 		cdi->mask &= ~CDC_DVD_R;
+	if (!(buf[8 + 4] & 0x01)) {
+		printk(KERN_INFO "ide-cd: audio play not advertised in caps "
+				"page for drive model [%s], enabling quirk.\n",
+				drive->id->model);

if IDE_CD_FLAG_PLAY_AUDIO_OK flag is not set the above message is _bogus_ (because the quirk won't be enabled)

[ how's about just deleting the whole printk() to preserve simplicity? ]

Quoted 4 lines
drivers/ide/ide-cd.c +1 -0
+	}
 	if ((buf[8 + 4] & 0x01) || (cd->cd_flags & IDE_CD_FLAG_PLAY_AUDIO_OK))
 		cdi->mask &= ~CDC_PLAY_AUDIO;
 
Quoted 7 lines
Quoted 5 lines

drivers/ide/ide-cd.c static const struct cd_list_entry ide_cd_quirks_list[] = { · line 1937

Quoted 5 lines
drivers/ide/ide-cd.c +1 -0
@@ -1929,6 +1934,7 @@ static const struct cd_list_entry ide_cd_quirks_list[] = {
 	{ "MATSHITADVD-ROM SR-8186", NULL,   IDE_CD_FLAG_PLAY_AUDIO_OK	    },
 	{ "MATSHITADVD-ROM SR-8176", NULL,   IDE_CD_FLAG_PLAY_AUDIO_OK	    },
 	{ "MATSHITADVD-ROM SR-8174", NULL,   IDE_CD_FLAG_PLAY_AUDIO_OK	    },
+	{ "Optiarc DVD RW AD-5200A", NULL,   IDE_CD_FLAG_PLAY_AUDIO_OK      },

But the flag _is_ set because the drive is added to the quirk list (line above ^) and

Quoted 5 lines
                                ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

(fortunately) majority of ATAPI drives are not in the quirk list ;-)

[ and they may have (buf[8 + 4] & 0x01) == 0 ]

Bart,

Quoted 7 lines

i think i probably didn't express myself clear enough before. The patch serves the sole purpose to show exactly whether buf[8 + 4] & 0x01) == 0 and is in no way the final solution. I just wanted to make sure the caps page doesn't set the audioplay bit. The final solution would be to add the drive to the quirks list only, i guess, in case you don't have a better idea.

Sorry for the misunderstanding :=).

Doh, all the time I thought that it is the final patch to be merged. :-)

[ I like to merge patches ASAP so please give some hint like putting
"Bart: DO NOT APPLY!" after "---" for draft/testing patches. Thanks. ]

#9 lore raw
In reply to Borislav Petkov
Quoted 66 lines · Borislav Petkov
Quoted 53 lines · Bartlomiej Zolnierkiewicz

Hi,

Quoted 26 lines · Borislav Petkov
Quoted 14 lines · Stefan Bader

Hello Borislav,

I worked on a problem with an DVD driver (model=Optiarc DVD RW AD-5200A) which obviously has the same problem as some Matshita drives. The following patch was reported to enabled audio playing on this drive. Would this approach be suitable for upstream or are there other solutions to this problem?

Regards,
Stefan

drivers/ide/ide-cd.c int ide_cdrom_probe_capabilities (ide_drive_t *drive) · line 2995

Quoted 9 lines
drivers/ide/ide-cd.c +2 -1
@@ -2988,7 +2988,8 @@ int ide_cdrom_probe_capabilities (ide_drive_t *drive)
 	if (strcmp(drive->id->model, "MATSHITADVD-ROM SR-8187") == 0 ||
 	    strcmp(drive->id->model, "MATSHITADVD-ROM SR-8186") == 0 ||
 	    strcmp(drive->id->model, "MATSHITADVD-ROM SR-8176") == 0 ||
-	    strcmp(drive->id->model, "MATSHITADVD-ROM SR-8174") == 0)
+	    strcmp(drive->id->model, "MATSHITADVD-ROM SR-8174") == 0 ||
+	    strcmp(drive->id->model, "Optiarc DVD RW AD-5200A") == 0)
 		CDROM_CONFIG_FLAGS(drive)->audio_play = 1;
 
 #if ! STANDARD_ATAPI

Hi Stefan,

Quoted 8 lines

just to make sure that the audioplay bit is not set in the capabilities page, can you please try the following patch applied against 2.6.25-rc2 and send me the output. Thanks!

@Bart: by the way, this cdi->mask thingy is kinda unintuitive doing double negation to check whether a feature is supported or not. Yeah, this comes from "above," i.e. uniform cdrom layer. But still, shouldn't we use a cdi->caps_flags or something similar instead, which mirrors the caps page bits setting?

It seems so (at least having negative flags is very unintuitive) but they might be some reason for this ugliness, Jens?

[ Please also remember that since cdrom layer is _uniform_ it may be not
possible and/or desirable to have 1-1 mapping between caps page bits and the future cdi->caps_flags. ]

Quoted 9 lines

commit 435f0f4496a1b32af2d542f43b2370a890fe2f83

Author: Borislav Petkov <[email protected]>
Date:   Sat Feb 16 09:56:36 2008 +0100

ide-cd: Enable audio play quirk for Optiarc DVD RW AD-5200A drive

Quoted 11 lines

drivers/ide/ide-cd.c int ide_cdrom_probe_capabilities (ide_drive_t *drive) · line 1753

Quoted 9 lines
drivers/ide/ide-cd.c +1 -0
index f77db6b..2c9d06e 100644
@@ -1750,6 +1750,10 @@ int ide_cdrom_probe_capabilities (ide_drive_t *drive)
 		cdi->mask &= ~(CDC_DVD_RAM | CDC_RAM);
 	if (buf[8 + 3] & 0x10)
 		cdi->mask &= ~CDC_DVD_R;
+	if (!(buf[8 + 4] & 0x01)) {

Hmm, shouldn't there be '&& (cd->cd_flags & IDE_CD_FLAG_PLAY_AUDIO_OK)' to prevent false positives?

Quoted 10 lines

I wanted to see whether the caps page reports the audioplay bit off...

Quoted 2 lines

drivers/ide/ide-cd.c int ide_cdrom_probe_capabilities (ide_drive_t *drive) · line 1754

Quoted 1 line
drivers/ide/ide-cd.c +1 -0
+		printk(KERN_INFO "ide-cd: audio play not advertised in caps page,"

Would be nice to also printk() the device name.

... but printing the device model is actually a good idea and this will rule out false positives, so Stefan, please drop the previous patch and test the updated one below. Thanks.

Hi Borislav,

the problem is that I don't own this drive myself and the owner is running a 2.6.22 kernel and is normally not doing any kernel compiles. But I could provide him a modified patch.
Though, if you just want to know whether the cap bit was really unset, I think we know this already. When I got the problem report we checked /proc/sys/dev/cdrom/info and that showed the "Can play audio" bit as 0. Which is the reason I gave the owner the patch for adding the model to the excemption list. And from his feedback I take that the drive plays audio tracks with the patch in use.

Stefan

Quoted 34 lines

commit 6cc44b0ce5c9270b15d456eb9ffa91b855e4e0d0

Author: Borislav Petkov <[email protected]>
Date:   Sat Feb 16 09:56:36 2008 +0100

ide-cd: Enable audio play quirk for Optiarc DVD RW AD-5200A drive

drivers/ide/ide-cd.c +6 -0
index f77db6b..4c9984f 100644
@@ -1750,6 +1750,11 @@ int ide_cdrom_probe_capabilities (ide_drive_t *drive)
 		cdi->mask &= ~(CDC_DVD_RAM | CDC_RAM);
 	if (buf[8 + 3] & 0x10)
 		cdi->mask &= ~CDC_DVD_R;
+	if (!(buf[8 + 4] & 0x01)) {
+		printk(KERN_INFO "ide-cd: audio play not advertised in caps "
+				"page for drive model [%s], enabling quirk.\n",
+				drive->id->model);
+	}
 	if ((buf[8 + 4] & 0x01) || (cd->cd_flags & IDE_CD_FLAG_PLAY_AUDIO_OK))
 		cdi->mask &= ~CDC_PLAY_AUDIO;
 
@@ -1929,6 +1934,7 @@ static const struct cd_list_entry ide_cd_quirks_list[] = {
 	{ "MATSHITADVD-ROM SR-8186", NULL,   IDE_CD_FLAG_PLAY_AUDIO_OK	    },
 	{ "MATSHITADVD-ROM SR-8176", NULL,   IDE_CD_FLAG_PLAY_AUDIO_OK	    },
 	{ "MATSHITADVD-ROM SR-8174", NULL,   IDE_CD_FLAG_PLAY_AUDIO_OK	    },
+	{ "Optiarc DVD RW AD-5200A", NULL,   IDE_CD_FLAG_PLAY_AUDIO_OK      },
 	{ NULL, NULL, 0 }
 };
 
#10 lore raw
Quoted 80 lines · Stefan Bader
Quoted 66 lines · Borislav Petkov
Quoted 53 lines · Bartlomiej Zolnierkiewicz

Hi,

Quoted 26 lines · Borislav Petkov
Quoted 14 lines · Stefan Bader

Hello Borislav,

I worked on a problem with an DVD driver (model=Optiarc DVD RW AD-5200A) which obviously has the same problem as some Matshita drives. The following patch was reported to enabled audio playing on this drive. Would this approach be suitable for upstream or are there other solutions to this problem?

Regards,
Stefan

drivers/ide/ide-cd.c int ide_cdrom_probe_capabilities (ide_drive_t *drive) · line 2995

Quoted 9 lines
drivers/ide/ide-cd.c +2 -1
@@ -2988,7 +2988,8 @@ int ide_cdrom_probe_capabilities (ide_drive_t *drive)
 	if (strcmp(drive->id->model, "MATSHITADVD-ROM SR-8187") == 0 ||
 	    strcmp(drive->id->model, "MATSHITADVD-ROM SR-8186") == 0 ||
 	    strcmp(drive->id->model, "MATSHITADVD-ROM SR-8176") == 0 ||
-	    strcmp(drive->id->model, "MATSHITADVD-ROM SR-8174") == 0)
+	    strcmp(drive->id->model, "MATSHITADVD-ROM SR-8174") == 0 ||
+	    strcmp(drive->id->model, "Optiarc DVD RW AD-5200A") == 0)
 		CDROM_CONFIG_FLAGS(drive)->audio_play = 1;
 
 #if ! STANDARD_ATAPI

Hi Stefan,

Quoted 8 lines

just to make sure that the audioplay bit is not set in the capabilities page, can you please try the following patch applied against 2.6.25-rc2 and send me the output. Thanks!

@Bart: by the way, this cdi->mask thingy is kinda unintuitive doing double negation to check whether a feature is supported or not. Yeah, this comes from "above," i.e. uniform cdrom layer. But still, shouldn't we use a cdi->caps_flags or something similar instead, which mirrors the caps page bits setting?

It seems so (at least having negative flags is very unintuitive) but they might be some reason for this ugliness, Jens?

[ Please also remember that since cdrom layer is _uniform_ it may be not
possible and/or desirable to have 1-1 mapping between caps page bits and the future cdi->caps_flags. ]

Quoted 9 lines

commit 435f0f4496a1b32af2d542f43b2370a890fe2f83

Author: Borislav Petkov <[email protected]>
Date:   Sat Feb 16 09:56:36 2008 +0100

ide-cd: Enable audio play quirk for Optiarc DVD RW AD-5200A drive

Quoted 11 lines

drivers/ide/ide-cd.c int ide_cdrom_probe_capabilities (ide_drive_t *drive) · line 1753

Quoted 9 lines
drivers/ide/ide-cd.c +1 -0
index f77db6b..2c9d06e 100644
@@ -1750,6 +1750,10 @@ int ide_cdrom_probe_capabilities (ide_drive_t *drive)
 		cdi->mask &= ~(CDC_DVD_RAM | CDC_RAM);
 	if (buf[8 + 3] & 0x10)
 		cdi->mask &= ~CDC_DVD_R;
+	if (!(buf[8 + 4] & 0x01)) {

Hmm, shouldn't there be '&& (cd->cd_flags & IDE_CD_FLAG_PLAY_AUDIO_OK)' to prevent false positives?

Quoted 10 lines

I wanted to see whether the caps page reports the audioplay bit off...

Quoted 2 lines

drivers/ide/ide-cd.c int ide_cdrom_probe_capabilities (ide_drive_t *drive) · line 1754

Quoted 1 line
drivers/ide/ide-cd.c +1 -0
+		printk(KERN_INFO "ide-cd: audio play not advertised in caps page,"

Would be nice to also printk() the device name.

... but printing the device model is actually a good idea and this will rule out false positives, so Stefan, please drop the previous patch and test the updated one below. Thanks.

Hi Borislav,

Quoted 9 lines

the problem is that I don't own this drive myself and the owner is running a 2.6.22 kernel and is normally not doing any kernel compiles. But I could provide him a modified patch.
Though, if you just want to know whether the cap bit was really unset, I think we know this already. When I got the problem report we checked /proc/sys/dev/cdrom/info and that showed the "Can play audio" bit as 0. Which is the reason I gave the owner the patch for adding the model to the excemption list. And from his feedback I take that the drive plays audio tracks with the patch in use.

Borislav, I guess that this is good enough proof that audioplay bit is off.

Could you please send me the final version of the patch?

Quoted 33 lines

Stefan

Quoted 31 lines

commit 6cc44b0ce5c9270b15d456eb9ffa91b855e4e0d0

Author: Borislav Petkov <[email protected]>
Date:   Sat Feb 16 09:56:36 2008 +0100

ide-cd: Enable audio play quirk for Optiarc DVD RW AD-5200A drive

drivers/ide/ide-cd.c +6 -0
index f77db6b..4c9984f 100644
@@ -1750,6 +1750,11 @@ int ide_cdrom_probe_capabilities (ide_drive_t *drive)
 		cdi->mask &= ~(CDC_DVD_RAM | CDC_RAM);
 	if (buf[8 + 3] & 0x10)
 		cdi->mask &= ~CDC_DVD_R;
+	if (!(buf[8 + 4] & 0x01)) {
+		printk(KERN_INFO "ide-cd: audio play not advertised in caps "
+				"page for drive model [%s], enabling quirk.\n",
+				drive->id->model);
+	}
 	if ((buf[8 + 4] & 0x01) || (cd->cd_flags & IDE_CD_FLAG_PLAY_AUDIO_OK))
 		cdi->mask &= ~CDC_PLAY_AUDIO;
 
@@ -1929,6 +1934,7 @@ static const struct cd_list_entry ide_cd_quirks_list[] = {
 	{ "MATSHITADVD-ROM SR-8186", NULL,   IDE_CD_FLAG_PLAY_AUDIO_OK	    },
 	{ "MATSHITADVD-ROM SR-8176", NULL,   IDE_CD_FLAG_PLAY_AUDIO_OK	    },
 	{ "MATSHITADVD-ROM SR-8174", NULL,   IDE_CD_FLAG_PLAY_AUDIO_OK	    },
+	{ "Optiarc DVD RW AD-5200A", NULL,   IDE_CD_FLAG_PLAY_AUDIO_OK      },
 	{ NULL, NULL, 0 }
#11 lore raw
Quoted 91 lines · Bartlomiej Zolnierkiewicz
Quoted 80 lines · Stefan Bader
Quoted 66 lines · Borislav Petkov
Quoted 53 lines · Bartlomiej Zolnierkiewicz

Hi,

Quoted 26 lines · Borislav Petkov
Quoted 14 lines · Stefan Bader

Hello Borislav,

I worked on a problem with an DVD driver (model=Optiarc DVD RW AD-5200A) which obviously has the same problem as some Matshita drives. The following patch was reported to enabled audio playing on this drive. Would this approach be suitable for upstream or are there other solutions to this problem?

Regards,
Stefan

drivers/ide/ide-cd.c int ide_cdrom_probe_capabilities (ide_drive_t *drive) · line 2995

Quoted 9 lines
drivers/ide/ide-cd.c +2 -1
@@ -2988,7 +2988,8 @@ int ide_cdrom_probe_capabilities (ide_drive_t *drive)
 	if (strcmp(drive->id->model, "MATSHITADVD-ROM SR-8187") == 0 ||
 	    strcmp(drive->id->model, "MATSHITADVD-ROM SR-8186") == 0 ||
 	    strcmp(drive->id->model, "MATSHITADVD-ROM SR-8176") == 0 ||
-	    strcmp(drive->id->model, "MATSHITADVD-ROM SR-8174") == 0)
+	    strcmp(drive->id->model, "MATSHITADVD-ROM SR-8174") == 0 ||
+	    strcmp(drive->id->model, "Optiarc DVD RW AD-5200A") == 0)
 		CDROM_CONFIG_FLAGS(drive)->audio_play = 1;
 
 #if ! STANDARD_ATAPI

Hi Stefan,

Quoted 8 lines

just to make sure that the audioplay bit is not set in the capabilities page, can you please try the following patch applied against 2.6.25-rc2 and send me the output. Thanks!

@Bart: by the way, this cdi->mask thingy is kinda unintuitive doing double negation to check whether a feature is supported or not. Yeah, this comes from "above," i.e. uniform cdrom layer. But still, shouldn't we use a cdi->caps_flags or something similar instead, which mirrors the caps page bits setting?

It seems so (at least having negative flags is very unintuitive) but they might be some reason for this ugliness, Jens?

[ Please also remember that since cdrom layer is _uniform_ it may be not
possible and/or desirable to have 1-1 mapping between caps page bits and the future cdi->caps_flags. ]

Quoted 9 lines

commit 435f0f4496a1b32af2d542f43b2370a890fe2f83

Author: Borislav Petkov <[email protected]>
Date:   Sat Feb 16 09:56:36 2008 +0100

ide-cd: Enable audio play quirk for Optiarc DVD RW AD-5200A drive

Quoted 11 lines

drivers/ide/ide-cd.c int ide_cdrom_probe_capabilities (ide_drive_t *drive) · line 1753

Quoted 9 lines
drivers/ide/ide-cd.c +1 -0
index f77db6b..2c9d06e 100644
@@ -1750,6 +1750,10 @@ int ide_cdrom_probe_capabilities (ide_drive_t *drive)
 		cdi->mask &= ~(CDC_DVD_RAM | CDC_RAM);
 	if (buf[8 + 3] & 0x10)
 		cdi->mask &= ~CDC_DVD_R;
+	if (!(buf[8 + 4] & 0x01)) {

Hmm, shouldn't there be '&& (cd->cd_flags & IDE_CD_FLAG_PLAY_AUDIO_OK)' to prevent false positives?

Quoted 10 lines

I wanted to see whether the caps page reports the audioplay bit off...

Quoted 2 lines

drivers/ide/ide-cd.c int ide_cdrom_probe_capabilities (ide_drive_t *drive) · line 1754

Quoted 1 line
drivers/ide/ide-cd.c +1 -0
+		printk(KERN_INFO "ide-cd: audio play not advertised in caps page,"

Would be nice to also printk() the device name.

... but printing the device model is actually a good idea and this will rule out false positives, so Stefan, please drop the previous patch and test the updated one below. Thanks.

Hi Borislav,

Quoted 9 lines

the problem is that I don't own this drive myself and the owner is running a 2.6.22 kernel and is normally not doing any kernel compiles. But I could provide him a modified patch.
Though, if you just want to know whether the cap bit was really unset, I think we know this already. When I got the problem report we checked /proc/sys/dev/cdrom/info and that showed the "Can play audio" bit as 0. Which is the reason I gave the owner the patch for adding the model to the excemption list. And from his feedback I take that the drive plays audio tracks with the patch in use.

Quoted 1 line

Borislav, I guess that this is good enough proof that audioplay bit is off.

indeed.

Quoted 1 line

Could you please send me the final version of the patch?

commit fa4af2fab0804bead4da6ecbf468118f05111229

Author: Borislav Petkov <[email protected]>
Date:   Sat Feb 16 09:56:36 2008 +0100

ide-cd: Enable audio play quirk for Optiarc DVD RW AD-5200A drive

drivers/ide/ide-cd.c +1 -0
index f77db6b..9e63c34 100644
@@ -1929,6 +1929,7 @@ static const struct cd_list_entry ide_cd_quirks_list[] = {
 	{ "MATSHITADVD-ROM SR-8186", NULL,   IDE_CD_FLAG_PLAY_AUDIO_OK	    },
 	{ "MATSHITADVD-ROM SR-8176", NULL,   IDE_CD_FLAG_PLAY_AUDIO_OK	    },
 	{ "MATSHITADVD-ROM SR-8174", NULL,   IDE_CD_FLAG_PLAY_AUDIO_OK	    },
+	{ "Optiarc DVD RW AD-5200A", NULL,   IDE_CD_FLAG_PLAY_AUDIO_OK      },
 	{ NULL, NULL, 0 }
 };
 
Signature · 2 lines
Regards/Gruß,
    Boris.