Skip to contentSkip to conversation

staging: sm750fb: constify fix_id array

PATCH Superseded Hungyu Lin 2 participants 4 messages Apr 1

Subsystems (3)

v3 of 4 Superseded Latest is v4 →

lore
Share & tools
lore

Hungyu Lin

author
#0 lore raw

PATCH v3 1/1 staging: sm750fb: constify fix_id array

Make the static fix_id array const-qualified so it can be placed in read-only memory.

---
Changes in v3:
- Drop g_fbmode change as it is modified at runtime.
- Resend as a new thread as requested.

---
 drivers/staging/sm750fb/sm750.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/staging/sm750fb/sm750.c +1 -1
index 9a42a08c8..f7b5b22f1 100644
@@ -728,7 +728,7 @@ static int lynxfb_set_fbinfo(struct fb_info *info, int index)
 		lynx750_ext, NULL, vesa_modes,
 	};
 	int cdb[] = {ARRAY_SIZE(lynx750_ext), 0, VESA_MODEDB_SIZE};
-	static const char *fix_id[2] = {
+	static const char * const fix_id[2] = {
 		"sm750_fb1", "sm750_fb2",
 	};
 
Signature · 1 line
2.34.1
#1 lore raw
Quoted 18 lines · Hungyu Lin

Make the static fix_id array const-qualified so it can be placed in read-only memory.

---
Changes in v3:
- Drop g_fbmode change as it is modified at runtime.
- Resend as a new thread as requested.

---
 drivers/staging/sm750fb/sm750.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/staging/sm750fb/sm750.c +0 -0
index 9a42a08c8..f7b5b22f1 100644

drivers/staging/sm750fb/sm750.c static int lynxfb_set_fbinfo(struct fb_info *info, int index) · line 731

Quoted 5 lines
drivers/staging/sm750fb/sm750.c +1 -1
@@ -728,7 +728,7 @@ static int lynxfb_set_fbinfo(struct fb_info *info, int index)
 		lynx750_ext, NULL, vesa_modes,
 	};
 	int cdb[] = {ARRAY_SIZE(lynx750_ext), 0, VESA_MODEDB_SIZE};
-	static const char *fix_id[2] = {
+	static const char * const fix_id[2] = {

Wait, is this really right?

You have a const char * array, and now you make the pointer to that const char * array const as well? But that was static already, so what exactly just got moved to read only memory? What does the compiler do differently here? How was this tested?

I need a bit more "proof" that this change really is correct as the const * and * to const and static stuff here always is confusing. There used to be some tool that would "unwind" this to figure it all out, but I can't find that at the moment...

thanks,

greg k-h

#2 lore raw
Quoted 32 lines · Greg Kroah-Hartman
Quoted 18 lines · Hungyu Lin

Make the static fix_id array const-qualified so it can be placed in read-only memory.

---
Changes in v3:
- Drop g_fbmode change as it is modified at runtime.
- Resend as a new thread as requested.

---
 drivers/staging/sm750fb/sm750.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/staging/sm750fb/sm750.c +0 -0
index 9a42a08c8..f7b5b22f1 100644

drivers/staging/sm750fb/sm750.c static int lynxfb_set_fbinfo(struct fb_info *info, int index) · line 731

Quoted 5 lines
drivers/staging/sm750fb/sm750.c +1 -1
@@ -728,7 +728,7 @@ static int lynxfb_set_fbinfo(struct fb_info *info, int index)
 		lynx750_ext, NULL, vesa_modes,
 	};
 	int cdb[] = {ARRAY_SIZE(lynx750_ext), 0, VESA_MODEDB_SIZE};
-	static const char *fix_id[2] = {
+	static const char * const fix_id[2] = {

Wait, is this really right?

You have a const char * array, and now you make the pointer to that const char * array const as well? But that was static already, so what exactly just got moved to read only memory? What does the compiler do differently here? How was this tested?

Quoted 4 lines

I need a bit more "proof" that this change really is correct as the const * and * to const and static stuff here always is confusing. There used to be some tool that would "unwind" this to figure it all out, but I can't find that at the moment...

Ah, found it:

	https://cdecl.org/

So:
const char * fix_id[2]
is:
declare fix_id as array 2 of pointer to const char

and:
const char * const fix_id[2]
is:
declare fix_id as array 2 of const pointer to const char

so I think that's right and your patch is ok?

But clarifiying what exactly got moved from where into where, would be good to have done. Please look at the output of the compiler for proof of that and put that in the changelog text for the next version.

thanks,

greg k-h

author
#3 lore raw

Hi Greg,

Thanks for the clarification.

I've addressed this in v4 by clarifying the const semantics and adding 'nm' output showing the symbol is placed in a read-only section.

Thanks,
Hungyu