Skip to contentSkip to conversation

nilfs2: Use __field_struct() for a bitwise field

PATCH In review Bart Van Assche 3 participants 5 messages May 6, 2024 – May 7, 2024

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

PATCH 1/1 nilfs2: Use __field_struct() for a bitwise field

As one can see in include/trace/stages/stage4_event_fields.h, the implementation of __field() uses the is_signed_type() macro. As one can see in commit dcf8e5633e2e ("tracing: Define the is_signed_type() macro once"), there has been an attempt to not make is_signed_type() trigger sparse warnings for bitwise types. Despite that change, sparse complains when passing a bitwise type to is_signed_type(). It is not clear to me why.

Follow the example of <trace/events/initcall.h> and suppress the following sparse warnings by changing __field() into __field_struct():

fs/nilfs2/segment.c: note: in included file (through
include/trace/trace_events.h, include/trace/define_trace.h, include/trace/events/nilfs2.h):
./include/trace/events/nilfs2.h:191:1: warning: cast to restricted
blk_opf_t
./include/trace/events/nilfs2.h:191:1: warning: restricted blk_opf_t
degrades to integer
./include/trace/events/nilfs2.h:191:1: warning: restricted blk_opf_t
degrades to integer

---
 include/trace/events/nilfs2.h | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)
include/trace/events/nilfs2.h +5 -1
index 8efc6236f57c..8880c11733dd 100644
@@ -200,7 +200,11 @@ TRACE_EVENT(nilfs2_mdt_submit_block,
 		    __field(struct inode *, inode)
 		    __field(unsigned long, ino)
 		    __field(unsigned long, blkoff)
-		    __field(enum req_op, mode)
+		    /*
+		     * Use field_struct() to avoid is_signed_type() on the
+		     * bitwise type enum req_op.
+		     */
+		    __field_struct(enum req_op, mode)
 	    ),
 
 	    TP_fast_assign(
maintainer
#1 lore raw
Quoted 38 lines · Bart Van Assche

As one can see in include/trace/stages/stage4_event_fields.h, the implementation of __field() uses the is_signed_type() macro. As one can see in commit dcf8e5633e2e ("tracing: Define the is_signed_type() macro once"), there has been an attempt to not make is_signed_type() trigger sparse warnings for bitwise types. Despite that change, sparse complains when passing a bitwise type to is_signed_type(). It is not clear to me why.

Follow the example of <trace/events/initcall.h> and suppress the following sparse warnings by changing __field() into __field_struct():

fs/nilfs2/segment.c: note: in included file (through
include/trace/trace_events.h, include/trace/define_trace.h, include/trace/events/nilfs2.h):
./include/trace/events/nilfs2.h:191:1: warning: cast to restricted
blk_opf_t
./include/trace/events/nilfs2.h:191:1: warning: restricted blk_opf_t
degrades to integer
./include/trace/events/nilfs2.h:191:1: warning: restricted blk_opf_t
degrades to integer

---
 include/trace/events/nilfs2.h | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)
include/trace/events/nilfs2.h +0 -0
index 8efc6236f57c..8880c11733dd 100644
@@ -200,7 +200,11 @@ TRACE_EVENT(nilfs2_mdt_submit_block,
                    __field(struct inode *, inode)
                    __field(unsigned long, ino)

include/trace/events/nilfs2.h TRACE_EVENT(nilfs2_mdt_submit_block, · line 210

Quoted 10 lines
include/trace/events/nilfs2.h +5 -1
                    __field(unsigned long, blkoff)
-                   __field(enum req_op, mode)
+                   /*
+                    * Use field_struct() to avoid is_signed_type() on the
+                    * bitwise type enum req_op.
+                    */
+                   __field_struct(enum req_op, mode)
            ),
 
            TP_fast_assign(

Bart, thank you very much for all your help.

This is a bit technical and may be debatable. But this can actually eliminate the sparse warnings.
So I'm thinking of having Andrew pick this up instead of my patch currently pending in the mm tree.

If anyone has any objections, please let us know.

Thanks,
Ryusuke Konishi

#3 lore raw
Quoted 3 lines · On Tue, 7 May 2024 05:55:27 -0700 Bart Van Assche <[email protected]> wrote:
Quoted 2 lines · Ryusuke Konishi

So I'm thinking of having Andrew pick this up instead of my patch currently pending in the mm tree.

Please identify "my patch"?

Quoted 5 lines

(+Andrew)

The patch Ryusuke is referring to is available here:
https://lore.kernel.org/linux-nilfs/[email protected]/

I'm not subscribed to linux-nilfs. It's good to cc linux-kernel on everything for this reason. Resend, please?

#4 lore raw
Quoted 7 lines · Andrew Morton
Quoted 3 lines · On Tue, 7 May 2024 05:55:27 -0700 Bart Van Assche <[email protected]> wrote:
Quoted 2 lines · Ryusuke Konishi

So I'm thinking of having Andrew pick this up instead of my patch currently pending in the mm tree.

Please identify "my patch"?

I think it's this patch from one week ago: "[PATCH -mm 1/2] nilfs2: use integer type instead of enum req_op for event tracing header" (https://lore.kernel.org/linux-nilfs/CAKFNMokwnkku4tHcaAJPmFaBONZqoTMOLtKh_A7kcjzoxj3QZw@mail.gmail.com/T/#mfbbd7edb4a3ecc2d819869e58fb5a460dcb5b4bd)

Thanks,

Bart.