Skip to contentSkip to conversation

kernel/audit.c: remove unused exports

PATCH In review Adrian Bunk 4 participants 10 messages Nov 28, 2006 – Dec 11, 2006 Audit

lore
Share & tools
lore

Adrian Bunk

author
#0 lore raw

PATCH 1/1 kernel/audit.c: remove unused exports

This patch removes the following unused EXPORT_SYMBOL's:
- audit_log_start
- audit_log_end
- audit_log_format
- audit_log

Signed-off-by Adrian Bunk <[email protected]>

---

This patch was already sent on:
- 20 Apr 2006

kernel/audit.c renamed from kernel/audit.c.old +0 -4
@@ -1092,7 +1092,3 @@
 	}
 }
 
-EXPORT_SYMBOL(audit_log_start);
-EXPORT_SYMBOL(audit_log_end);
-EXPORT_SYMBOL(audit_log_format);
-EXPORT_SYMBOL(audit_log);
#1 lore raw
Quoted 6 lines · Adrian Bunk

This patch removes the following unused EXPORT_SYMBOL's:
- audit_log_start
- audit_log_end
- audit_log_format
- audit_log

Quoted 6 lines
Signed-off-by Adrian Bunk <[email protected]>

---

This patch was already sent on:
- 20 Apr 2006

I recall the opinion at the time was that it was considered useful to allow third party modules to generate audit messages. Has anything changed?

Tony

#2 lore raw
Quoted 2 lines · Tony Jones

I recall the opinion at the time was that it was considered useful to allow third party modules to generate audit messages. Has anything changed?

It should at least be EXPORT_SYMBOL_GPL, not EXPORT_SYMBOL.

Signature · 1 line
dwmw2
#3 lore raw
Quoted 5 lines · David Woodhouse
Quoted 2 lines · Tony Jones

I recall the opinion at the time was that it was considered useful to allow third party modules to generate audit messages. Has anything changed?

It should at least be EXPORT_SYMBOL_GPL, not EXPORT_SYMBOL.

Agreed. That sounds fine.

Tony

author
#4 lore raw

PATCH 1/1 kernel/audit.c: change the exports to EXPORT_SYMBOL_GPL

In reply to David Woodhouse
Quoted 5 lines · David Woodhouse
Quoted 2 lines · Tony Jones

I recall the opinion at the time was that it was considered useful to allow third party modules to generate audit messages. Has anything changed?

It should at least be EXPORT_SYMBOL_GPL, not EXPORT_SYMBOL.

So let's do at least the minimal patch noone disagrees with.

Quoted 1 line

dwmw2

cu
Adrian

<-- snip -->

This patch changes some completely unused audit exports from EXPORT_SYMBOL to EXPORT_SYMBOL_GPL.

They are still completely unused, but hopefully some of the theoretical code that might use it will appear in the kernel in the near future...

Signed-off-by Adrian Bunk <[email protected]>
kernel/audit.c renamed from kernel/audit.c.old +4 -4
@@ -1209,7 +1209,7 @@
 	}
 }
 
-EXPORT_SYMBOL(audit_log_start);
-EXPORT_SYMBOL(audit_log_end);
-EXPORT_SYMBOL(audit_log_format);
-EXPORT_SYMBOL(audit_log);
+EXPORT_SYMBOL_GPL(audit_log_start);
+EXPORT_SYMBOL_GPL(audit_log_end);
+EXPORT_SYMBOL_GPL(audit_log_format);
+EXPORT_SYMBOL_GPL(audit_log);
#5 lore raw

Re: [2.6 patch] kernel/audit.c: change the exports to EXPORT_SYMBOL_GPL

Quoted 3 lines · Adrian Bunk
Quoted 1 line

It should at least be EXPORT_SYMBOL_GPL, not EXPORT_SYMBOL.

So let's do at least the minimal patch noone disagrees with.

This looks fine to me.

-Steve

author
#6 lore raw
In reply to Tony Jones
Quoted 7 lines · Tony Jones
Quoted 6 lines · Adrian Bunk

This patch removes the following unused EXPORT_SYMBOL's:
- audit_log_start
- audit_log_end
- audit_log_format
- audit_log

Quoted 8 lines
Quoted 6 lines
Signed-off-by Adrian Bunk <[email protected]>

---

This patch was already sent on:
- 20 Apr 2006

I recall the opinion at the time was that it was considered useful to allow

"it was considered" = "you were the one person who said this"

Quoted 1 line

third party modules to generate audit messages. Has anything changed?

Why are these third party modules you are talking about not yet merged into the kernel and when will they be submitted for inclusion?

Quoted 1 line

Tony

cu
Adrian

Signature · 4 lines
       "Is there not promise of rain?" Ling Tan asked suddenly out
        of the darkness. There had been need of rain for many days.
       "Only a promise," Lao Er said.
                                       Pearl S. Buck - Dragon Seed
#7 lore raw
Quoted 3 lines · Adrian Bunk
Quoted 1 line

I recall the opinion at the time was that it was considered useful to allow

"it was considered" = "you were the one person who said this"

Actually I believe Steve Grubb also said it so the above is factually incorrect :)

Quoted 4 lines
Quoted 1 line

third party modules to generate audit messages. Has anything changed?

Why are these third party modules you are talking about not yet merged into the kernel and when will they be submitted for inclusion?

We're talking about basic logging functionality here. I know you love to swing the cleanup hammer but ....... you get the idea.

Tony

#8 lore raw
Quoted 2 lines · Tony Jones

Actually I believe Steve Grubb also said it so the above is factually incorrect :)

Yes, I did. I thought this was a dead issue since there was no follow up discussion when I said it.

Quoted 4 lines
Quoted 4 lines
Quoted 1 line

third party modules to generate audit messages.  Has anything changed?

Why are these third party modules you are talking about not yet merged into the kernel and when will they be submitted for inclusion?

It would appear that a new audit patch to ipsec wants these and a couple more things. So, at this point we should have something queued in netdev tree that will use these.

Quoted 1 line

We're talking about basic logging functionality here.

Exactly. I can see 3rd party security vendors perhaps using this interface as well. I don't personally know of any that are using it, nor am I in discussions with any, but I could see them eventually finding this and using it as they use other basic interfaces like syslog or file access.

-Steve

author
#9 lore raw
Quoted 10 lines · Steve Grubb

On Wednesday 29 November 2006 02:42, Tony Jones wrote:
...

Quoted 4 lines
Quoted 4 lines
Quoted 1 line

third party modules to generate audit messages.  Has anything changed?

Why are these third party modules you are talking about not yet merged into the kernel and when will they be submitted for inclusion?

It would appear that a new audit patch to ipsec wants these and a couple more things. So, at this point we should have something queued in netdev tree that will use these.

AFAIR, this ipsec patch only adds non-modular users.

Quoted 6 lines
Quoted 1 line

We're talking about basic logging functionality here.

Exactly. I can see 3rd party security vendors perhaps using this interface as well. I don't personally know of any that are using it, nor am I in discussions with any, but I could see them eventually finding this and using it as they use other basic interfaces like syslog or file access.

Reality has teached me that the only way to find out whether unused code is really unused is to send a patch and listen whether someone is screaming because he has an actual user. Then get this patch merged and listen whether someone is screaming because he had an actual user.

As an example, it has just turned out that at about half of the drivers that were not compiling and marked as BROKEN since at least 2.6.0 had patches floating around somewhere and sending a patch to remove code actually brought this code back into life.

We are talking about exports that are unused since they were introduced. Either there are no users, or the existing users need the biggest possible cluebat (removing the exports) reminding them that they should get their ass up and submit their modules for inclusion in the kernel.

Re-adding them when users are merged into the kernel will always be trivial.

Quoted 1 line

-Steve

cu
Adrian

Signature · 4 lines
       "Is there not promise of rain?" Ling Tan asked suddenly out
        of the darkness. There had been need of rain for many days.
       "Only a promise," Lao Er said.
                                       Pearl S. Buck - Dragon Seed