[alsa-devel] [PATCH] ALSA: fix building with CONFIG_SND_PROC_FS disabled
A recent change to split the procfs support for ALSA turned out to be incomplete, leading to build errors:
sound/core/seq/seq_info.c:61:12: error: redefinition of 'snd_seq_info_init' sound/core/seq/seq_device.c: In function 'seq_dev_proc_init': sound/core/seq/seq_device.c:276:2: error: 'info_entry' undeclared (first use in this function) sound/core/seq/seq_device.c: In function 'alsa_seq_device_exit': sound/core/seq/seq_device.c:309:22: error: 'info_entry' undeclared (first use in this function)
This replaces some more instances of CONFIG_PROC_FS with CONFIG_SND_PROC_FS, as the original patch started. It is enough to avoid all randconfig regressions I saw on ARM because of the CONFIG_SND_PROC_FS introduction.
Signed-off-by: Arnd Bergmann arnd@arndb.de Fixes: cd6a65036f0e ("ALSA: replace CONFIG_PROC_FS with CONFIG_SND_PROC_FS") ---
I assume this was already reported. If not, please apply my patch or fold it into the original commit.
diff --git a/sound/core/seq/Makefile b/sound/core/seq/Makefile index b29ffe835205..b65fa5a1943b 100644 --- a/sound/core/seq/Makefile +++ b/sound/core/seq/Makefile @@ -7,7 +7,7 @@ snd-seq-device-objs := seq_device.o snd-seq-objs := seq.o seq_lock.o seq_clientmgr.o seq_memory.o seq_queue.o \ seq_fifo.o seq_prioq.o seq_timer.o \ seq_system.o seq_ports.o -snd-seq-$(CONFIG_PROC_FS) += seq_info.o +snd-seq-$(CONFIG_SND_PROC_FS) += seq_info.o snd-seq-midi-objs := seq_midi.o snd-seq-midi-emul-objs := seq_midi_emul.o snd-seq-midi-event-objs := seq_midi_event.o diff --git a/sound/core/seq/seq_device.c b/sound/core/seq/seq_device.c index 288af56a43cd..c4acf17e9f5e 100644 --- a/sound/core/seq/seq_device.c +++ b/sound/core/seq/seq_device.c @@ -272,7 +272,7 @@ EXPORT_SYMBOL_GPL(snd_seq_driver_unregister);
static int __init seq_dev_proc_init(void) { -#ifdef CONFIG_PROC_FS +#ifdef CONFIG_SND_PROC_FS info_entry = snd_info_create_module_entry(THIS_MODULE, "drivers", snd_seq_root); if (info_entry == NULL) @@ -305,7 +305,7 @@ static void __exit alsa_seq_device_exit(void) #ifdef CONFIG_MODULES cancel_work_sync(&autoload_work); #endif -#ifdef CONFIG_PROC_FS +#ifdef CONFIG_SND_PROC_FS snd_info_free_entry(info_entry); #endif bus_unregister(&snd_seq_bus_type);
-----Original Message----- From: Arnd Bergmann [mailto:arnd@arndb.de] Sent: Friday, May 29, 2015 4:47 PM To: Takashi Iwai Cc: broonie@kernel.org; alsa-devel@alsa-project.org; linux-arm- kernel@lists.infradead.org; Jie, Yang Subject: [PATCH] ALSA: fix building with CONFIG_SND_PROC_FS disabled
A recent change to split the procfs support for ALSA turned out to be incomplete, leading to build errors:
sound/core/seq/seq_info.c:61:12: error: redefinition of 'snd_seq_info_init' sound/core/seq/seq_device.c: In function 'seq_dev_proc_init': sound/core/seq/seq_device.c:276:2: error: 'info_entry' undeclared (first use in this function) sound/core/seq/seq_device.c: In function 'alsa_seq_device_exit': sound/core/seq/seq_device.c:309:22: error: 'info_entry' undeclared (first use in this function)
This replaces some more instances of CONFIG_PROC_FS with CONFIG_SND_PROC_FS, as the original patch started. It is enough to avoid all randconfig regressions I saw on ARM because of the CONFIG_SND_PROC_FS introduction.
Signed-off-by: Arnd Bergmann arnd@arndb.de Fixes: cd6a65036f0e ("ALSA: replace CONFIG_PROC_FS with CONFIG_SND_PROC_FS")
I assume this was already reported. If not, please apply my patch or fold it into the original commit.
Thank you, Seems this already fixed by Takashi's latest series?
~Keyon
diff --git a/sound/core/seq/Makefile b/sound/core/seq/Makefile index b29ffe835205..b65fa5a1943b 100644 --- a/sound/core/seq/Makefile +++ b/sound/core/seq/Makefile @@ -7,7 +7,7 @@ snd-seq-device-objs := seq_device.o snd-seq-objs := seq.o seq_lock.o seq_clientmgr.o seq_memory.o seq_queue.o \ seq_fifo.o seq_prioq.o seq_timer.o \ seq_system.o seq_ports.o -snd-seq-$(CONFIG_PROC_FS) += seq_info.o +snd-seq-$(CONFIG_SND_PROC_FS) += seq_info.o snd-seq-midi-objs := seq_midi.o snd-seq-midi-emul-objs := seq_midi_emul.o snd-seq-midi-event-objs := seq_midi_event.o diff --git a/sound/core/seq/seq_device.c b/sound/core/seq/seq_device.c index 288af56a43cd..c4acf17e9f5e 100644 --- a/sound/core/seq/seq_device.c +++ b/sound/core/seq/seq_device.c @@ -272,7 +272,7 @@ EXPORT_SYMBOL_GPL(snd_seq_driver_unregister);
static int __init seq_dev_proc_init(void) { -#ifdef CONFIG_PROC_FS +#ifdef CONFIG_SND_PROC_FS info_entry = snd_info_create_module_entry(THIS_MODULE, "drivers", snd_seq_root); if (info_entry == NULL) @@ -305,7 +305,7 @@ static void __exit alsa_seq_device_exit(void) #ifdef CONFIG_MODULES cancel_work_sync(&autoload_work); #endif -#ifdef CONFIG_PROC_FS +#ifdef CONFIG_SND_PROC_FS snd_info_free_entry(info_entry); #endif bus_unregister(&snd_seq_bus_type);
At Fri, 29 May 2015 09:03:35 +0000, Jie, Yang wrote:
-----Original Message----- From: Arnd Bergmann [mailto:arnd@arndb.de] Sent: Friday, May 29, 2015 4:47 PM To: Takashi Iwai Cc: broonie@kernel.org; alsa-devel@alsa-project.org; linux-arm- kernel@lists.infradead.org; Jie, Yang Subject: [PATCH] ALSA: fix building with CONFIG_SND_PROC_FS disabled
A recent change to split the procfs support for ALSA turned out to be incomplete, leading to build errors:
sound/core/seq/seq_info.c:61:12: error: redefinition of 'snd_seq_info_init' sound/core/seq/seq_device.c: In function 'seq_dev_proc_init': sound/core/seq/seq_device.c:276:2: error: 'info_entry' undeclared (first use in this function) sound/core/seq/seq_device.c: In function 'alsa_seq_device_exit': sound/core/seq/seq_device.c:309:22: error: 'info_entry' undeclared (first use in this function)
This replaces some more instances of CONFIG_PROC_FS with CONFIG_SND_PROC_FS, as the original patch started. It is enough to avoid all randconfig regressions I saw on ARM because of the CONFIG_SND_PROC_FS introduction.
Signed-off-by: Arnd Bergmann arnd@arndb.de Fixes: cd6a65036f0e ("ALSA: replace CONFIG_PROC_FS with CONFIG_SND_PROC_FS")
I assume this was already reported. If not, please apply my patch or fold it into the original commit.
Thank you, Seems this already fixed by Takashi's latest series?
Yes.
Takashi
~Keyon
diff --git a/sound/core/seq/Makefile b/sound/core/seq/Makefile index b29ffe835205..b65fa5a1943b 100644 --- a/sound/core/seq/Makefile +++ b/sound/core/seq/Makefile @@ -7,7 +7,7 @@ snd-seq-device-objs := seq_device.o snd-seq-objs := seq.o seq_lock.o seq_clientmgr.o seq_memory.o seq_queue.o \ seq_fifo.o seq_prioq.o seq_timer.o \ seq_system.o seq_ports.o -snd-seq-$(CONFIG_PROC_FS) += seq_info.o +snd-seq-$(CONFIG_SND_PROC_FS) += seq_info.o snd-seq-midi-objs := seq_midi.o snd-seq-midi-emul-objs := seq_midi_emul.o snd-seq-midi-event-objs := seq_midi_event.o diff --git a/sound/core/seq/seq_device.c b/sound/core/seq/seq_device.c index 288af56a43cd..c4acf17e9f5e 100644 --- a/sound/core/seq/seq_device.c +++ b/sound/core/seq/seq_device.c @@ -272,7 +272,7 @@ EXPORT_SYMBOL_GPL(snd_seq_driver_unregister);
static int __init seq_dev_proc_init(void) { -#ifdef CONFIG_PROC_FS +#ifdef CONFIG_SND_PROC_FS info_entry = snd_info_create_module_entry(THIS_MODULE, "drivers", snd_seq_root); if (info_entry == NULL) @@ -305,7 +305,7 @@ static void __exit alsa_seq_device_exit(void) #ifdef CONFIG_MODULES cancel_work_sync(&autoload_work); #endif -#ifdef CONFIG_PROC_FS +#ifdef CONFIG_SND_PROC_FS snd_info_free_entry(info_entry); #endif bus_unregister(&snd_seq_bus_type);
participants (3)
-
Arnd Bergmann -
Jie, Yang -
Takashi Iwai