[PATCH v6 0/2] Fix compiler warnings/errors in SUNRPC and MACB

[PATCH v6 0/2] Fix compiler warnings/errors in SUNRPC and MACB

[Sean Chang]

This series addresses compiler warnings and a build error identified across 
different architectures and configurations.

The first patch refactors the dfprintk macros in the SUNRPC subsystem. This 
global fix silences unused variable warnings and, by incorporating 
no_printk(), ensures continued compiler type checking. It also includes 
a specific fix for a build error in nfsd (nfsfh.c) uncovered by syzbot 
in non-debug configurations.

The second patch fixes a format-truncation warning in the MACB ethernet
driver by ensuring the snprintf output fits within the destination buffer.

v6:
- Instead of wrapping dprintk in #if blocks, completely remove the 
  redundant RPC_IFDEBUG() macro and associated #if guards in nfsfh.c and 
  svc_rdma_transport.c.
- Remove the #if guard around the static helper nlmdbg_cookie2a() in 
  svclock.c.
- Verified via .lst and nm that both variables (buf) and helper functions 
  (nlmdbg_cookie2a) are fully optimized out by the compiler when 
  CONFIG_SUNRPC_DEBUG is disabled.

v5:
- Simplify dfprintk and dfprintk_rcu macros by removing the redundant 'fmt' 
  argument and calling no_printk(__VA_ARGS__) directly, as suggested by 
  David Laight.
- Fix a build error in fs/nfsd/nfsfh.c reported by syzbot. The error was 
  caused by a mismatch between the variable's lifecycle (defined via 
  RPC_IFDEBUG) and its usage in dprintk.
- Add Link tag to the syzbot build failure report.
- Update series title to reflect the general nature of the fixes.

v4:
- Refactor patch 1 to use no_printk() in sunrpc headers instead of marking
  variables as __maybe_unused. This provides a cleaner, global fix and
  enables compiler type checking.

v3:
- Expand commit descriptions to include technical details regarding macro
  expansion, as requested by Andrew Lunn.
- Test the different platform, such as ARM, ARM64, X86_64.

v2:
- Split the original treewide patch into subsystem-specific commits.
- Added more detailed commit descriptions to satisfy checkpatch.

Sean Chang (2):
  sunrpc: simplify dprintk macros and cleanup redundant debug guards
  net: macb: use ethtool_sprintf to fill ethtool stats strings

 drivers/net/ethernet/cadence/macb_main.c | 7 ++-----
 fs/lockd/svclock.c                       | 2 --
 fs/nfsd/nfsfh.c                          | 2 +-
 include/linux/sunrpc/debug.h             | 6 ++----
 net/sunrpc/xprtrdma/svc_rdma_transport.c | 4 +---
 5 files changed, 6 insertions(+), 15 deletions(-)

-- 
2.34.1

[PATCH v6 1/2] sunrpc: simplify dprintk macros and cleanup redundant debug guards

[Sean Chang]

When CONFIG_SUNRPC_DEBUG is disabled, the dfprintk() macros currently
expand to empty do-while loops. This causes variables used solely
within these calls to appear unused, triggering -Wunused-variable
warnings.

Following David Laight's suggestion, simplify the macro definitions by
removing the unnecessary 'fmt' argument and using no_printk(__VA_ARGS__)
directly. This ensures the compiler performs type checking and "sees"
the variables, silencing the warnings without emitting any code.

Verification with .lst files under -O2 confirms that the compiler
successfully performs "dead code elimination". Even when variables
(like char buf[] in nfsfh.c) or static helper functions (like
nlmdbg_cookie2a in svclock.c) are declared without #ifdef, they are
completely optimized out (no stack allocation, no symbol references in
the final executable) as they are only referenced within no_printk().

This allows for significant cleanup:
- Remove RPC_IFDEBUG() and associated #if blocks in fs/nfsd/nfsfh.c
  and net/sunrpc/xprtrdma/svc_rdma_transport.c.
- Remove the #if IS_ENABLED(CONFIG_SUNRPC_DEBUG) guard around
  nlmdbg_cookie2a in fs/lockd/svclock.c.
- Consolidate the dprintk definition to be more idiomatic.

This fixes the build errors reported by the kernel test robot while
improving code maintainability.

Link: https://lore.kernel.org/all/69a2e269.050a0220.3a55be.003e.GAE@google.com/
Reported-by: kernel test robot <lkp@intel.com>
Closes: https://lore.kernel.org/oe-kbuild-all/202603010612.uRmHYMsi-lkp@intel.com/
Closes: https://lore.kernel.org/oe-kbuild-all/202603010852.3RKXCwyF-lkp@intel.com/
Closes: https://lore.kernel.org/oe-kbuild-all/202603010808.w3TtG6fC-lkp@intel.com/
Suggested-by: Andrew Lunn <andrew@lunn.ch>
Suggested-by: David Laight <david.laight.linux@gmail.com>
Signed-off-by: Sean Chang <seanwascoding@gmail.com>
---
 fs/lockd/svclock.c                       | 2 --
 fs/nfsd/nfsfh.c                          | 2 +-
 include/linux/sunrpc/debug.h             | 6 ++----
 net/sunrpc/xprtrdma/svc_rdma_transport.c | 4 +---
 4 files changed, 4 insertions(+), 10 deletions(-)

diff --git a/fs/lockd/svclock.c b/fs/lockd/svclock.c
index 255a847ca0b6..4bac32f009f6 100644
--- a/fs/lockd/svclock.c
+++ b/fs/lockd/svclock.c
@@ -53,7 +53,6 @@ static const struct rpc_call_ops nlmsvc_grant_ops;
 static LIST_HEAD(nlm_blocked);
 static DEFINE_SPINLOCK(nlm_blocked_lock);
 
-#if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
 static const char *nlmdbg_cookie2a(const struct nlm_cookie *cookie)
 {
 	/*
@@ -80,7 +79,6 @@ static const char *nlmdbg_cookie2a(const struct nlm_cookie *cookie)
 
 	return buf;
 }
-#endif
 
 /*
  * Insert a blocked lock into the global list
diff --git a/fs/nfsd/nfsfh.c b/fs/nfsd/nfsfh.c
index ed85dd43da18..cb3fb91d18c0 100644
--- a/fs/nfsd/nfsfh.c
+++ b/fs/nfsd/nfsfh.c
@@ -105,7 +105,7 @@ static __be32 nfsd_setuser_and_check_port(struct svc_rqst *rqstp,
 {
 	/* Check if the request originated from a secure port. */
 	if (rqstp && !nfsd_originating_port_ok(rqstp, cred, exp)) {
-		RPC_IFDEBUG(char buf[RPC_MAX_ADDRBUFLEN]);
+		char buf[RPC_MAX_ADDRBUFLEN];
 		dprintk("nfsd: request from insecure port %s!\n",
 		        svc_print_addr(rqstp, buf, sizeof(buf)));
 		return nfserr_perm;
diff --git a/include/linux/sunrpc/debug.h b/include/linux/sunrpc/debug.h
index eb4bd62df319..f6f2a106eeaf 100644
--- a/include/linux/sunrpc/debug.h
+++ b/include/linux/sunrpc/debug.h
@@ -49,12 +49,10 @@ do {									\
 	}								\
 } while (0)
 
-# define RPC_IFDEBUG(x)		x
 #else
 # define ifdebug(fac)		if (0)
-# define dfprintk(fac, fmt, ...)	do {} while (0)
-# define dfprintk_rcu(fac, fmt, ...)	do {} while (0)
-# define RPC_IFDEBUG(x)
+# define dfprintk(fac, ...)		no_printk(__VA_ARGS__)
+# define dfprintk_rcu(fac, ...)	no_printk(__VA_ARGS__)
 #endif
 
 /*
diff --git a/net/sunrpc/xprtrdma/svc_rdma_transport.c b/net/sunrpc/xprtrdma/svc_rdma_transport.c
index 9b623849723e..ba6fe0fd387d 100644
--- a/net/sunrpc/xprtrdma/svc_rdma_transport.c
+++ b/net/sunrpc/xprtrdma/svc_rdma_transport.c
@@ -414,7 +414,7 @@ static struct svc_xprt *svc_rdma_accept(struct svc_xprt *xprt)
 	struct ib_qp_init_attr qp_attr;
 	struct ib_device *dev;
 	int ret = 0;
-	RPC_IFDEBUG(struct sockaddr *sap);
+	struct sockaddr *sap;
 
 	listen_rdma = container_of(xprt, struct svcxprt_rdma, sc_xprt);
 	clear_bit(XPT_CONN, &xprt->xpt_flags);
@@ -560,7 +560,6 @@ static struct svc_xprt *svc_rdma_accept(struct svc_xprt *xprt)
 		goto errout;
 	}
 
-#if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
 	dprintk("svcrdma: new connection accepted on device %s:\n", dev->name);
 	sap = (struct sockaddr *)&newxprt->sc_cm_id->route.addr.src_addr;
 	dprintk("    local address   : %pIS:%u\n", sap, rpc_get_port(sap));
@@ -571,7 +570,6 @@ static struct svc_xprt *svc_rdma_accept(struct svc_xprt *xprt)
 	dprintk("    rdma_rw_ctxs    : %d\n", ctxts);
 	dprintk("    max_requests    : %d\n", newxprt->sc_max_requests);
 	dprintk("    ord             : %d\n", conn_param.initiator_depth);
-#endif
 
 	return &newxprt->sc_xprt;
 
-- 
2.34.1

Re: [PATCH v6 1/2] sunrpc: simplify dprintk macros and cleanup redundant debug guards

[Andy Shevchenko]

On Mon, Mar 02, 2026 at 12:17:08AM +0800, Sean Chang wrote:
> When CONFIG_SUNRPC_DEBUG is disabled, the dfprintk() macros currently
> expand to empty do-while loops. This causes variables used solely
> within these calls to appear unused, triggering -Wunused-variable
> warnings.
> 
> Following David Laight's suggestion, simplify the macro definitions by
> removing the unnecessary 'fmt' argument and using no_printk(__VA_ARGS__)
> directly. This ensures the compiler performs type checking and "sees"
> the variables, silencing the warnings without emitting any code.
> 
> Verification with .lst files under -O2 confirms that the compiler
> successfully performs "dead code elimination". Even when variables
> (like char buf[] in nfsfh.c) or static helper functions (like
> nlmdbg_cookie2a in svclock.c) are declared without #ifdef, they are
> completely optimized out (no stack allocation, no symbol references in
> the final executable) as they are only referenced within no_printk().
> 
> This allows for significant cleanup:
> - Remove RPC_IFDEBUG() and associated #if blocks in fs/nfsd/nfsfh.c
>   and net/sunrpc/xprtrdma/svc_rdma_transport.c.
> - Remove the #if IS_ENABLED(CONFIG_SUNRPC_DEBUG) guard around
>   nlmdbg_cookie2a in fs/lockd/svclock.c.
> - Consolidate the dprintk definition to be more idiomatic.
> 
> This fixes the build errors reported by the kernel test robot while
> improving code maintainability.

If you feel something has to be done against nfsd-next, please rebase and send
as a standalone patch.

-- 
With Best Regards,
Andy Shevchenko

[PATCH v6 2/2] net: macb: use ethtool_sprintf to fill ethtool stats strings

[Sean Chang]

The RISC-V toolchain triggers a stringop-truncation warning when using
snprintf() with a fixed ETH_GSTRING_LEN (32 bytes) buffer.

Convert the driver to use the modern ethtool_sprintf() API from
linux/ethtool.h. This removes the need for manual snprintf() and
memcpy() calls, handles the 32-byte padding automatically, and
simplifies the logic by removing manual pointer arithmetic.

Suggested-by: Andrew Lunn <andrew@lunn.ch>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: Sean Chang <seanwascoding@gmail.com>
---
 drivers/net/ethernet/cadence/macb_main.c | 7 ++-----
 1 file changed, 2 insertions(+), 5 deletions(-)

diff --git a/drivers/net/ethernet/cadence/macb_main.c b/drivers/net/ethernet/cadence/macb_main.c
index 5bc35f651ebd..79ca19097b2d 100644
--- a/drivers/net/ethernet/cadence/macb_main.c
+++ b/drivers/net/ethernet/cadence/macb_main.c
@@ -3145,7 +3145,6 @@ static int gem_get_sset_count(struct net_device *dev, int sset)
 
 static void gem_get_ethtool_strings(struct net_device *dev, u32 sset, u8 *p)
 {
-	char stat_string[ETH_GSTRING_LEN];
 	struct macb *bp = netdev_priv(dev);
 	struct macb_queue *queue;
 	unsigned int i;
@@ -3158,10 +3157,8 @@ static void gem_get_ethtool_strings(struct net_device *dev, u32 sset, u8 *p)
 			       ETH_GSTRING_LEN);
 
 		for (q = 0, queue = bp->queues; q < bp->num_queues; ++q, ++queue) {
-			for (i = 0; i < QUEUE_STATS_LEN; i++, p += ETH_GSTRING_LEN) {
-				snprintf(stat_string, ETH_GSTRING_LEN, "q%d_%s",
-						q, queue_statistics[i].stat_string);
-				memcpy(p, stat_string, ETH_GSTRING_LEN);
+			for (i = 0; i < QUEUE_STATS_LEN; i++) {
+				ethtool_sprintf(&p, "q%u_%s", q, queue_statistics[i].stat_string);
 			}
 		}
 		break;
-- 
2.34.1

Re: [PATCH v6 0/2] Fix compiler warnings/errors in SUNRPC and MACB

[Andrew Lunn]

> - Verified via .lst and nm that both variables (buf) and helper functions 
>   (nlmdbg_cookie2a) are fully optimized out by the compiler when 
>   CONFIG_SUNRPC_DEBUG is disabled.

Thanks for doing this bit. It was needed to show my guess was correct.

Reviewed-by: Andrew Lunn <andrew@lunn.ch>

    Andrew

Re: [PATCH v6 0/2] Fix compiler warnings/errors in SUNRPC and MACB

[Andy Shevchenko]

On Sun, Mar 01, 2026 at 06:39:44PM +0100, Andrew Lunn wrote:
> > - Verified via .lst and nm that both variables (buf) and helper functions 
> >   (nlmdbg_cookie2a) are fully optimized out by the compiler when 
> >   CONFIG_SUNRPC_DEBUG is disabled.
> 
> Thanks for doing this bit. It was needed to show my guess was correct.
> 
> Reviewed-by: Andrew Lunn <andrew@lunn.ch>

It already has your tag. My Q in v5 still stands: Why is this a series and not
standalone changes?

Also I just gave a tag, please split these two and send individually.

-- 
With Best Regards,
Andy Shevchenko

Re: [PATCH v6 0/2] Fix compiler warnings/errors in SUNRPC and MACB

[Sean Chang]

On Mon, Mar 2, 2026 at 9:26 PM Andy Shevchenko
<andriy.shevchenko@intel.com> wrote:
>
> On Sun, Mar 01, 2026 at 06:39:44PM +0100, Andrew Lunn wrote:
> > > - Verified via .lst and nm that both variables (buf) and helper functions
> > >   (nlmdbg_cookie2a) are fully optimized out by the compiler when
> > >   CONFIG_SUNRPC_DEBUG is disabled.
> >
> > Thanks for doing this bit. It was needed to show my guess was correct.
> >
> > Reviewed-by: Andrew Lunn <andrew@lunn.ch>
>
> It already has your tag. My Q in v5 still stands: Why is this a series and not
> standalone changes?
>
> Also I just gave a tag, please split these two and send individually.
>

Hi Andy,
Thanks for the suggestion. I have split these into two standalone patches
as you requested:
1. The SUNRPC/NFSD macro simplification [1]
2. The Ethtool/macb redundant guard cleanup [2]

I've also verified the optimization again via .lst files to ensure everything
is clean.

[1] https://lore.kernel.org/all/20260302142931.49108-1-seanwascoding@gmail.com/
[2] https://lore.kernel.org/all/20260302161818.63651-1-seanwascoding@gmail.com/
---
Best Regards,
Sean