Skip to contentSkip to conversation

Rename SECTOR_SIZE to IDE_SECTOR_SIZE

RFC In review Chris Wedgwood 6 participants 9 messages Oct 27, 2004

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

RFC 1/1 Rename SECTOR_SIZE to IDE_SECTOR_SIZE

The token SECTOR_SIZE is used in multiple places that have (almost) the same defintion everywhere.

How do people feel about rename this vague token?

===== drivers/char/ftape/lowlevel/ftape-bsm.h 1.2 vs edited =====

drivers/char/ftape/lowlevel/ftape-bsm.h +1 -1
@@ -36,7 +36,7 @@
 
 /*  maximum (format code 4) bad sector map size (bytes).
  */
-#define BAD_SECTOR_MAP_SIZE     (29 * SECTOR_SIZE - 256)
+#define BAD_SECTOR_MAP_SIZE     (29 * IDE_SECTOR_SIZE - 256)
 
 /*  format code 4 bad sector entry, ftape uses this
  *  internally for all format codes

===== drivers/ide/ide-cd.c 1.96 vs edited =====

drivers/ide/ide-cd.c +27 -27

177 lines not on this page · raw message

===== drivers/ide/ide-cd.h 1.11 vs edited =====

drivers/ide/ide-cd.h +6 -9
@@ -43,14 +43,11 @@
 
 /************************************************************************/
 
-#define SECTOR_BITS 		9
-#ifndef SECTOR_SIZE
-#define SECTOR_SIZE		(1 << SECTOR_BITS)
-#endif
-#define SECTORS_PER_FRAME	(CD_FRAMESIZE >> SECTOR_BITS)
-#define SECTOR_BUFFER_SIZE	(CD_FRAMESIZE * 32)
-#define SECTORS_BUFFER		(SECTOR_BUFFER_SIZE >> SECTOR_BITS)
-#define SECTORS_MAX		(131072 >> SECTOR_BITS)
+#define CD_SECTOR_BITS 		9
+#define CD_SECTORS_PER_FRAME	(CD_FRAMESIZE >> CD_SECTOR_BITS)
+#define CD_SECTOR_BUFFER_SIZE	(CD_FRAMESIZE * 32)
+#define CD_SECTORS_BUFFER	(SECTOR_BUFFER_SIZE >> CD_SECTOR_BITS)
+#define CD_SECTORS_MAX		(131072 >> CD_SECTOR_BITS)
 
 #define BLOCKS_PER_FRAME	(CD_FRAMESIZE / BLOCK_SIZE)
 
@@ -84,7 +81,7 @@ struct ide_cd_config_flags {
 	__u8 supp_disc_present	: 1; /* Changer can report exact contents
 					of slots. */
 	__u8 limit_nframes	: 1; /* Drive does not provide data in
-					multiples of SECTOR_SIZE when more
+					multiples of CD_SECTOR_SIZE when more
 					than one interrupt is needed. */
 	__u8 seeking		: 1; /* Seeking in progress */
 	__u8 audio_play		: 1; /* can do audio related commands */

===== drivers/ide/ide-io.c 1.34 vs edited =====

drivers/ide/ide-io.c +1 -1
@@ -686,7 +686,7 @@ void ide_map_sg(ide_drive_t *drive, stru
 	if ((rq->flags & REQ_DRIVE_TASKFILE) == 0) {
 		hwif->sg_nents = blk_rq_map_sg(drive->queue, rq, sg);
 	} else {
-		sg_init_one(sg, rq->buffer, rq->nr_sectors * SECTOR_SIZE);
+		sg_init_one(sg, rq->buffer, rq->nr_sectors * IDE_SECTOR_SIZE);
 		hwif->sg_nents = 1;
 	}
 }

===== drivers/ide/ide-taskfile.c 1.69 vs edited =====

drivers/ide/ide-taskfile.c +3 -3
@@ -288,12 +288,12 @@ static void ide_pio_sector(ide_drive_t *
 	local_irq_save(flags);
 #endif
 	buf = kmap_atomic(page, KM_BIO_SRC_IRQ) +
-	      sg[hwif->cursg].offset + (hwif->cursg_ofs * SECTOR_SIZE);
+	      sg[hwif->cursg].offset + (hwif->cursg_ofs * IDE_SECTOR_SIZE);
 
 	hwif->nleft--;
 	hwif->cursg_ofs++;
 
-	if ((hwif->cursg_ofs * SECTOR_SIZE) == sg[hwif->cursg].length) {
+	if ((hwif->cursg_ofs * IDE_SECTOR_SIZE) == sg[hwif->cursg].length) {
 		hwif->cursg++;
 		hwif->cursg_ofs = 0;
 	}
@@ -488,7 +488,7 @@ int ide_diag_taskfile (ide_drive_t *driv
 		if (data_size == 0)
 			rq.nr_sectors = (args->hobRegister[IDE_NSECTOR_OFFSET] << 8) | args->tfRegister[IDE_NSECTOR_OFFSET];
 		else
-			rq.nr_sectors = data_size / SECTOR_SIZE;
+			rq.nr_sectors = data_size / IDE_SECTOR_SIZE;
 
 		if (!rq.nr_sectors) {
 			printk(KERN_ERR "%s: in/out command without data\n",

===== drivers/ide/pci/siimage.c 1.35 vs edited =====

drivers/ide/pci/siimage.c +1 -1
@@ -491,7 +491,7 @@ static int siimage_mmio_ide_dma_count (i
 #ifdef SIIMAGE_VIRTUAL_DMAPIO
 	struct request *rq	= HWGROUP(drive)->rq;
 	ide_hwif_t *hwif	= HWIF(drive);
-	u32 count		= (rq->nr_sectors * SECTOR_SIZE);
+	u32 count		= (rq->nr_sectors * IDE_SECTOR_SIZE);
 	u32 rcount		= 0;
 	unsigned long addr	= siimage_selreg(hwif, 0x1C);
 

===== include/linux/ide.h 1.148 vs edited =====

include/linux/ide.h +2 -2
@@ -202,8 +202,8 @@ typedef unsigned char	byte;	/* used ever
 #define PARTN_BITS	6	/* number of minor dev bits for partitions */
 #define PARTN_MASK	((1<<PARTN_BITS)-1)	/* a useful bit mask */
 #define MAX_DRIVES	2	/* per interface; 2 assumed by lots of code */
-#define SECTOR_SIZE	512
-#define SECTOR_WORDS	(SECTOR_SIZE / 4)	/* number of 32bit words per sector */
+#define IDE_SECTOR_SIZE	512
+#define SECTOR_WORDS	(IDE_SECTOR_SIZE / 4)	/* number of 32bit words per sector */
 #define IDE_LARGE_SEEK(b1,b2,t)	(((b1) > (b2) + (t)) || ((b2) > (b1) + (t)))
 
 /*
#1 lore raw
Quoted 4 lines · Chris Wedgwood

The token SECTOR_SIZE is used in multiple places that have (almost) the same defintion everywhere.

How do people feel about rename this vague token?

It's highly silly to rename the same name + the same value to multiple different names.

Put it in a common header somewhere, and only rename the oddballs (if any).

Jeff

#2 lore raw
Quoted 2 lines · Jeff Garzik

It's highly silly to rename the same name + the same value to multiple different names.

initially i was going to do that, but when i looked at the code i realized the problem is some of the users seem to be semantically different and potentially might want to be changed separate to the others

Quoted 2 lines

Put it in a common header somewhere, and only rename the oddballs (if any).

we could have UNIX_SECTOR_SIZE in blkdev.h but as i said, some users really are 512 for different reasons that might change (?)

#3 lore raw
Quoted 10 lines · Chris Wedgwood
Quoted 2 lines · Jeff Garzik

It's highly silly to rename the same name + the same value to multiple different names.

initially i was going to do that, but when i looked at the code i realized the problem is some of the users seem to be semantically different and potentially might want to be changed separate to the others

Quoted 5 lines
Quoted 2 lines

Put it in a common header somewhere, and only rename the oddballs (if any).

we could have UNIX_SECTOR_SIZE in blkdev.h but as i said, some users really are 512 for different reasons that might change (?)

Please, just call it IO_SECTOR_SIZE or BIO_SECTOR_SIZE to signify that it's the block io sector size. And use that where it applies, leave the rest of the unrelated cases alone.

Signature · 1 line
Jens Axboe
#4 lore raw

PATCH 1/1 Rename SECTOR_SIZE to BIO_SECTOR_SIZE

Rename (one of the uses of) SECTOR_SIZE to BIO_SECTOR_SIZE which is more appropriate.

---

Rediffed against the BK tree from a few minutes ago.

 drivers/char/ftape/lowlevel/ftape-bsm.h |    2 -
 drivers/ide/ide-cd.c                    |   54 ++++++++++++++++----------------
 drivers/ide/ide-cd.h                    |   15 +++-----
 drivers/ide/ide-io.c                    |    2 -
 drivers/ide/ide-taskfile.c              |    6 +--
 drivers/ide/pci/siimage.c               |    2 -
 include/linux/ide.h                     |    4 +-
 7 files changed, 41 insertions(+), 44 deletions(-)
drivers/char/ftape/lowlevel/ftape-bsm.h +1 -1
Index: cw-current/drivers/char/ftape/lowlevel/ftape-bsm.h
===================================================================
@@ -36,7 +36,7 @@
 
 /*  maximum (format code 4) bad sector map size (bytes).
  */
-#define BAD_SECTOR_MAP_SIZE     (29 * SECTOR_SIZE - 256)
+#define BAD_SECTOR_MAP_SIZE     (29 * BIO_SECTOR_SIZE - 256)
 
 /*  format code 4 bad sector entry, ftape uses this
  *  internally for all format codes
drivers/ide/ide-cd.c +27 -27

177 lines not on this page · raw message

drivers/ide/ide-cd.h +6 -9
Index: cw-current/drivers/ide/ide-cd.h
===================================================================
@@ -43,14 +43,11 @@
 
 /************************************************************************/
 
-#define SECTOR_BITS 		9
-#ifndef SECTOR_SIZE
-#define SECTOR_SIZE		(1 << SECTOR_BITS)
-#endif
-#define SECTORS_PER_FRAME	(CD_FRAMESIZE >> SECTOR_BITS)
-#define SECTOR_BUFFER_SIZE	(CD_FRAMESIZE * 32)
-#define SECTORS_BUFFER		(SECTOR_BUFFER_SIZE >> SECTOR_BITS)
-#define SECTORS_MAX		(131072 >> SECTOR_BITS)
+#define CD_SECTOR_BITS 		9
+#define CD_SECTORS_PER_FRAME	(CD_FRAMESIZE >> CD_SECTOR_BITS)
+#define CD_SECTOR_BUFFER_SIZE	(CD_FRAMESIZE * 32)
+#define CD_SECTORS_BUFFER	(SECTOR_BUFFER_SIZE >> CD_SECTOR_BITS)
+#define CD_SECTORS_MAX		(131072 >> CD_SECTOR_BITS)
 
 #define BLOCKS_PER_FRAME	(CD_FRAMESIZE / BLOCK_SIZE)
 
@@ -84,7 +81,7 @@
 	__u8 supp_disc_present	: 1; /* Changer can report exact contents
 					of slots. */
 	__u8 limit_nframes	: 1; /* Drive does not provide data in
-					multiples of SECTOR_SIZE when more
+					multiples of CD_SECTOR_SIZE when more
 					than one interrupt is needed. */
 	__u8 seeking		: 1; /* Seeking in progress */
 	__u8 audio_play		: 1; /* can do audio related commands */
drivers/ide/ide-io.c +1 -1
Index: cw-current/drivers/ide/ide-io.c
===================================================================
@@ -686,7 +686,7 @@
 	if ((rq->flags & REQ_DRIVE_TASKFILE) == 0) {
 		hwif->sg_nents = blk_rq_map_sg(drive->queue, rq, sg);
 	} else {
-		sg_init_one(sg, rq->buffer, rq->nr_sectors * SECTOR_SIZE);
+		sg_init_one(sg, rq->buffer, rq->nr_sectors * BIO_SECTOR_SIZE);
 		hwif->sg_nents = 1;
 	}
 }
drivers/ide/ide-taskfile.c +3 -3
Index: cw-current/drivers/ide/ide-taskfile.c
===================================================================
@@ -281,12 +281,12 @@
 	local_irq_save(flags);
 #endif
 	buf = kmap_atomic(page, KM_BIO_SRC_IRQ) +
-	      sg[hwif->cursg].offset + (hwif->cursg_ofs * SECTOR_SIZE);
+	      sg[hwif->cursg].offset + (hwif->cursg_ofs * BIO_SECTOR_SIZE);
 
 	hwif->nleft--;
 	hwif->cursg_ofs++;
 
-	if ((hwif->cursg_ofs * SECTOR_SIZE) == sg[hwif->cursg].length) {
+	if ((hwif->cursg_ofs * BIO_SECTOR_SIZE) == sg[hwif->cursg].length) {
 		hwif->cursg++;
 		hwif->cursg_ofs = 0;
 	}
@@ -481,7 +481,7 @@
 		if (data_size == 0)
 			rq.nr_sectors = (args->hobRegister[IDE_NSECTOR_OFFSET] << 8) | args->tfRegister[IDE_NSECTOR_OFFSET];
 		else
-			rq.nr_sectors = data_size / SECTOR_SIZE;
+			rq.nr_sectors = data_size / BIO_SECTOR_SIZE;
 
 		if (!rq.nr_sectors) {
 			printk(KERN_ERR "%s: in/out command without data\n",
drivers/ide/pci/siimage.c +1 -1
Index: cw-current/drivers/ide/pci/siimage.c
===================================================================
@@ -470,7 +470,7 @@
 #ifdef SIIMAGE_VIRTUAL_DMAPIO
 	struct request *rq	= HWGROUP(drive)->rq;
 	ide_hwif_t *hwif	= HWIF(drive);
-	u32 count		= (rq->nr_sectors * SECTOR_SIZE);
+	u32 count		= (rq->nr_sectors * BIO_SECTOR_SIZE);
 	u32 rcount		= 0;
 	unsigned long addr	= siimage_selreg(hwif, 0x1C);
 
include/linux/ide.h +2 -2
Index: cw-current/include/linux/ide.h
===================================================================
@@ -202,8 +202,8 @@
 #define PARTN_BITS	6	/* number of minor dev bits for partitions */
 #define PARTN_MASK	((1<<PARTN_BITS)-1)	/* a useful bit mask */
 #define MAX_DRIVES	2	/* per interface; 2 assumed by lots of code */
-#define SECTOR_SIZE	512
-#define SECTOR_WORDS	(SECTOR_SIZE / 4)	/* number of 32bit words per sector */
+#define BIO_SECTOR_SIZE	512
+#define SECTOR_WORDS	(BIO_SECTOR_SIZE / 4)	/* number of 32bit words per sector */
 #define IDE_LARGE_SEEK(b1,b2,t)	(((b1) > (b2) + (t)) || ((b2) > (b1) + (t)))
 
 /*
#5 lore raw

Re: [PATCH] Rename SECTOR_SIZE to BIO_SECTOR_SIZE

Quoted 8 lines · Chris Wedgwood

Rename (one of the uses of) SECTOR_SIZE to BIO_SECTOR_SIZE which is more appropriate.

Index: cw-current/include/linux/ide.h
===================================================================

include/linux/ide.h line 207

Quoted 8 lines
include/linux/ide.h +2 -2
@@ -202,8 +202,8 @@
 #define PARTN_BITS	6	/* number of minor dev bits for partitions */
 #define PARTN_MASK	((1<<PARTN_BITS)-1)	/* a useful bit mask */
 #define MAX_DRIVES	2	/* per interface; 2 assumed by lots of code */
-#define SECTOR_SIZE	512
-#define SECTOR_WORDS	(SECTOR_SIZE / 4)	/* number of 32bit words per sector */
+#define BIO_SECTOR_SIZE	512
+#define SECTOR_WORDS	(BIO_SECTOR_SIZE / 4)	/* number of 32bit words per sector */
 #define IDE_LARGE_SEEK(b1,b2,t)	(((b1) > (b2) + (t)) || ((b2) > (b1) + (t)))

So shouldn't this be in bio.h now?

Signature · 1 line
Mathematics is the supreme nostalgia of our time.
#6 lore raw

Re: [PATCH] Rename SECTOR_SIZE to BIO_SECTOR_SIZE

Quoted 1 line · Matt Mackall

So shouldn't this be in bio.h now?

in a sense, nobody else but IDE uses it, and i just noticed SECTOR_WORDS is pretty much pointless so i can rediff fixing that too...

#7 lore raw

Re: [PATCH] Rename SECTOR_SIZE to BIO_SECTOR_SIZE

Quoted 5 lines · Chris Wedgwood
Quoted 1 line · Matt Mackall

So shouldn't this be in bio.h now?

in a sense, nobody else but IDE uses it, and i just noticed

Then whole rename doesn't make sense to me... sorry... :)

Quoted 2 lines

SECTOR_WORDS is pretty much pointless so i can rediff fixing that too...

SECTOR_WORDS can die

#8 lore raw

Re: [PATCH] Rename SECTOR_SIZE to BIO_SECTOR_SIZE

In reply to Matt Mackall
Quoted 4 lines · Matt Mackall
Quoted 2 lines · Chris Wedgwood

Rename (one of the uses of) SECTOR_SIZE to BIO_SECTOR_SIZE which is more appropriate.

Quoted 16 lines

include/linux/ide.h line 207

Quoted 13 lines

Index: cw-current/include/linux/ide.h
===================================================================

include/linux/ide.h +2 -2
@@ -202,8 +202,8 @@
 #define PARTN_BITS	6	/* number of minor dev bits for partitions */
 #define PARTN_MASK	((1<<PARTN_BITS)-1)	/* a useful bit mask */
 #define MAX_DRIVES	2	/* per interface; 2 assumed by lots of code */
-#define SECTOR_SIZE	512
-#define SECTOR_WORDS	(SECTOR_SIZE / 4)	/* number of 32bit words per sector */
+#define BIO_SECTOR_SIZE	512
+#define SECTOR_WORDS	(BIO_SECTOR_SIZE / 4)	/* number of 32bit words per sector */
 #define IDE_LARGE_SEEK(b1,b2,t)	(((b1) > (b2) + (t)) || ((b2) > (b1) + (t)))

So shouldn't this be in bio.h now?

And for uniformity, SECTOR_WORDS -> BIO_SECTOR_WORDS ?

--
J.A. Magallon <jamagallon()able!es> \ Software is like sex:
werewolf!able!es \ It's better when it's free
Mandrakelinux release 10.1 (Community) for i586
Linux 2.6.9-jam1 (gcc 3.4.1 (Mandrakelinux 10.1 3.4.1-4mdk)) #6