Skip to contentSkip to conversation

fix DocBook build

PATCH In review Jeff Garzik 2 participants 5 messages Sep 7, 2005

Subsystems (4)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 fix DocBook build

Please pull from 'upstream' branch of rsync://rsync.kernel.org/pub/scm/linux/kernel/git/jgarzik/misc-2.6.git

to obtain the following DocBook change:

 Documentation/DocBook/mcabook.tmpl |    2 +-
 drivers/net/wan/syncppp.c          |    1 +
 drivers/scsi/libata-core.c         |    4 ++--
 include/sound/pcm.h                |    2 +-
 kernel/sched.c                     |    1 +
 5 files changed, 6 insertions(+), 4 deletions(-)

commit 344babaa9d39b10b85cadec4e5335d43b52b4ec0

Author: Jeff Garzik <[email protected]>
Date:   Wed Sep 7 01:15:17 2005 -0400

[kernel-doc] fix various DocBook build problems/warnings

Most serious is fixing include/sound/pcm.h, which breaks the DocBook build.

The other stuff is just filling in things that cause warnings.

Documentation/DocBook/mcabook.tmpl +1 -1
@@ -96,7 +96,7 @@
 
   <chapter id="pubfunctions">
      <title>Public Functions Provided</title>
-!Earch/i386/kernel/mca.c
+!Edrivers/mca/mca-legacy.c
   </chapter>
 
   <chapter id="dmafunctions">
drivers/net/wan/syncppp.c +1 -0
@@ -1440,6 +1440,7 @@ static void sppp_print_bytes (u_char *p,
  *	@skb:	The buffer to process
  *	@dev:	The device it arrived on
  *	@p: Unused
+ *	@orig_dev: Unused
  *
  *	Protocol glue. This drives the deferred processing mode the poorer
  *	cards use. This can be called directly by cards that do not have
drivers/scsi/libata-core.c +2 -2
@@ -2531,7 +2531,7 @@ void swap_buf_le16(u16 *buf, unsigned in
  *	@ap: port to read/write
  *	@buf: data buffer
  *	@buflen: buffer length
- *	@do_write: read/write
+ *	@write_data: read/write
  *
  *	Transfer data from/to the device data register by MMIO.
  *
@@ -2577,7 +2577,7 @@ static void ata_mmio_data_xfer(struct at
  *	@ap: port to read/write
  *	@buf: data buffer
  *	@buflen: buffer length
- *	@do_write: read/write
+ *	@write_data: read/write
  *
  *	Transfer data from/to the device data register by PIO.
  *
include/sound/pcm.h +1 -1
@@ -903,7 +903,7 @@ int snd_pcm_format_unsigned(snd_pcm_form
 int snd_pcm_format_linear(snd_pcm_format_t format);
 int snd_pcm_format_little_endian(snd_pcm_format_t format);
 int snd_pcm_format_big_endian(snd_pcm_format_t format);
-/**
+/*
  * snd_pcm_format_cpu_endian - Check the PCM format is CPU-endian
  * @format: the format to check
  *
kernel/sched.c +1 -0
@@ -1478,6 +1478,7 @@ static inline void prepare_task_switch(r
 
 /**
  * finish_task_switch - clean up after a task-switch
+ * @rq: runqueue associated with task-switch
  * @prev: the thread we just switched away from.
  *
  * finish_task_switch must be called after the context switch, paired
#1 lore raw

At Wed, 7 Sep 2005 01:17:20 -0400,

Quoted 6 lines · Jeff Garzik

[kernel-doc] fix various DocBook build problems/warnings

Most serious is fixing include/sound/pcm.h, which breaks the DocBook build.

What is the error exactly? IIRC, it did work in the early version. Well, I need to test it by myself...

Takashi

#2 lore raw

At Wed, 07 Sep 2005 12:00:31 +0200,

Quoted 5 lines · I wrote:

At Wed, 7 Sep 2005 01:17:20 -0400,

Quoted 2 lines · Jeff Garzik
Quoted 7 lines
Quoted 4 lines

[kernel-doc] fix various DocBook build problems/warnings

Most serious is fixing include/sound/pcm.h, which breaks the DocBook build.

What is the error exactly? IIRC, it did work in the early version. Well, I need to test it by myself...

OK, I see it, too. How about the fix below, instead?

[PATCH] Fix DocBook build in sound/pcm.h

include/sound/pcm.h +1 -1
@@ -910,7 +910,7 @@ int snd_pcm_format_big_endian(snd_pcm_fo
  * Returns 1 if the given PCM format is CPU-endian, 0 if
  * opposite, or a negative error code if endian not specified.
  */
-/* int snd_pcm_format_cpu_endian(snd_pcm_format_t format); */
+int snd_pcm_format_cpu_endian(snd_pcm_format_t format);
 #ifdef SNDRV_LITTLE_ENDIAN
 #define snd_pcm_format_cpu_endian	snd_pcm_format_little_endian
 #else
#3 lore raw
Quoted 28 lines · Takashi Iwai

At Wed, 07 Sep 2005 12:00:31 +0200,

Quoted 11 lines · I wrote:

At Wed, 7 Sep 2005 01:17:20 -0400,

Quoted 5 lines · Jeff Garzik

[kernel-doc] fix various DocBook build problems/warnings

Most serious is fixing include/sound/pcm.h, which breaks the DocBook build.

What is the error exactly? IIRC, it did work in the early version. Well, I need to test it by myself...

OK, I see it, too. How about the fix below, instead?

[PATCH] Fix DocBook build in sound/pcm.h

include/sound/pcm.h int snd_pcm_format_big_endian(snd_pcm_fo · line 916

Quoted 8 lines
include/sound/pcm.h +1 -1
@@ -910,7 +910,7 @@ int snd_pcm_format_big_endian(snd_pcm_fo
  * Returns 1 if the given PCM format is CPU-endian, 0 if
  * opposite, or a negative error code if endian not specified.
  */
-/* int snd_pcm_format_cpu_endian(snd_pcm_format_t format); */
+int snd_pcm_format_cpu_endian(snd_pcm_format_t format);
 #ifdef SNDRV_LITTLE_ENDIAN
 #define snd_pcm_format_cpu_endian	snd_pcm_format_little_endian
 #else

I considered that, but decided not to, given the C pre-processor games that are occurring with that symbol shortly after the commented-out block.

Jeff

#4 lore raw

At Wed, 07 Sep 2005 07:27:54 -0400,

Quoted 26 lines · Jeff Garzik
Quoted 24 lines · Takashi Iwai

At Wed, 07 Sep 2005 12:00:31 +0200,

Quoted 11 lines · I wrote:

At Wed, 7 Sep 2005 01:17:20 -0400,

Quoted 5 lines · Jeff Garzik

[kernel-doc] fix various DocBook build problems/warnings

Most serious is fixing include/sound/pcm.h, which breaks the DocBook build.

What is the error exactly? IIRC, it did work in the early version. Well, I need to test it by myself...

OK, I see it, too. How about the fix below, instead?

[PATCH] Fix DocBook build in sound/pcm.h

Quoted 15 lines

include/sound/pcm.h int snd_pcm_format_big_endian(snd_pcm_fo · line 916

Quoted 12 lines
include/sound/pcm.h +1 -1
@@ -910,7 +910,7 @@ int snd_pcm_format_big_endian(snd_pcm_fo
  * Returns 1 if the given PCM format is CPU-endian, 0 if
  * opposite, or a negative error code if endian not specified.
  */
-/* int snd_pcm_format_cpu_endian(snd_pcm_format_t format); */
+int snd_pcm_format_cpu_endian(snd_pcm_format_t format);
 #ifdef SNDRV_LITTLE_ENDIAN
 #define snd_pcm_format_cpu_endian	snd_pcm_format_little_endian
 #else

I considered that, but decided not to, given the C pre-processor games that are occurring with that symbol shortly after the commented-out block.

The function doesn't exist but is declared just for docbook. Since it's replaced with the macro anyway, it shouldn't hurt.

To be safer, we can put #if 0 around the comment block and the pseudo function definition.

Takashi

include/sound/pcm.h +3 -1
@@ -903,6 +903,7 @@ int snd_pcm_format_unsigned(snd_pcm_form
 int snd_pcm_format_linear(snd_pcm_format_t format);
 int snd_pcm_format_little_endian(snd_pcm_format_t format);
 int snd_pcm_format_big_endian(snd_pcm_format_t format);
+#if 0 /* just for DocBook */
 /**
  * snd_pcm_format_cpu_endian - Check the PCM format is CPU-endian
  * @format: the format to check
@@ -910,7 +911,8 @@ int snd_pcm_format_big_endian(snd_pcm_fo
  * Returns 1 if the given PCM format is CPU-endian, 0 if
  * opposite, or a negative error code if endian not specified.
  */
-/* int snd_pcm_format_cpu_endian(snd_pcm_format_t format); */
+int snd_pcm_format_cpu_endian(snd_pcm_format_t format);
+#endif /* DocBook */
 #ifdef SNDRV_LITTLE_ENDIAN
 #define snd_pcm_format_cpu_endian	snd_pcm_format_little_endian
 #else