* [PATCH net-next v2] tg3: Propagate PHY reset retry failures in tg3_phy_reset_5703_4_5()
@ 2026-11-22 4:53 Ian Rogers
0 siblings, 0 replies; only message in thread
From: Ian Rogers @ 2026-11-22 4:53 UTC (permalink / raw)
To: Pavan Chebbi, Michael Chan, Andrew Lunn, David S. Miller,
Eric Dumazet, Jakub Kicinski, Paolo Abeni
Cc: netdev, linux-kernel, llvm, Elkin Cruz, Ian Rogers
In tg3_phy_reset_5703_4_5(), the 10-iteration retry loop discards the
return value of both tg3_readphy() calls on continue and never checks
err after the loop terminates. If MDIO reads fail across all 10
retries, phy9_orig is never initialized before being written back to
MII_CTRL1000; if tg3_phy_write_and_check_testpat() fails across all 10
retries, the function silently continues as if the PHY test pattern
succeeded.
Record err on both tg3_readphy() calls and return err if the retry
loop finishes with an error so dead MDIO hardware fails probe cleanly
via tg3_phy_probe(), and zero-initialize phy9_orig since Clang's
value-insensitive -Wconditional-uninitialized check does not correlate
the earlier continue on err != 0 with the post-loop 'if (err)' check.
Found via Clang -Wconditional-uninitialized on an x86_64 allyesconfig
build with AI coding assistant help. Verified by building
drivers/net/ethernet/broadcom/tg3.o with LLVM=1 W=1 (-Werror) and
-Wconditional-uninitialized; not runtime-tested on BCM5703/5704/5705
hardware.
Signed-off-by: Ian Rogers <irogers@google.com>
---
Notes:
v2:
- Target net-next and drop the Fixes tag (Andrew Lunn).
- Also zero-initialize phy9_orig so Clang's value-insensitive
-Wconditional-uninitialized check sees no uninitialized CFG path across
the earlier 'continue'.
v1: https://lore.kernel.org/r/20261119171021.1535172-1-irogers@google.com
drivers/net/ethernet/broadcom/tg3.c | 11 ++++++++---
1 file changed, 8 insertions(+), 3 deletions(-)
diff --git a/drivers/net/ethernet/broadcom/tg3.c b/drivers/net/ethernet/broadcom/tg3.c
index 95827df4cf6a..4d375328621c 100644
--- a/drivers/net/ethernet/broadcom/tg3.c
+++ b/drivers/net/ethernet/broadcom/tg3.c
@@ -2544,7 +2544,7 @@ static int tg3_phy_reset_chanpat(struct tg3 *tp)
static int tg3_phy_reset_5703_4_5(struct tg3 *tp)
{
- u32 reg32, phy9_orig;
+ u32 reg32, phy9_orig = 0;
int retries, do_phy_reset, err;
retries = 10;
@@ -2558,7 +2558,8 @@ static int tg3_phy_reset_5703_4_5(struct tg3 *tp)
}
/* Disable transmitter and interrupt. */
- if (tg3_readphy(tp, MII_TG3_EXT_CTRL, ®32))
+ err = tg3_readphy(tp, MII_TG3_EXT_CTRL, ®32);
+ if (err)
continue;
reg32 |= 0x3000;
@@ -2569,7 +2570,8 @@ static int tg3_phy_reset_5703_4_5(struct tg3 *tp)
BMCR_FULLDPLX | BMCR_SPEED1000);
/* Set to master mode. */
- if (tg3_readphy(tp, MII_CTRL1000, &phy9_orig))
+ err = tg3_readphy(tp, MII_CTRL1000, &phy9_orig);
+ if (err)
continue;
tg3_writephy(tp, MII_CTRL1000,
@@ -2587,6 +2589,9 @@ static int tg3_phy_reset_5703_4_5(struct tg3 *tp)
break;
} while (--retries);
+ if (err)
+ return err;
+
err = tg3_phy_reset_chanpat(tp);
if (err)
return err;
--
2.56.1.750.gf70dc682cf-goog
^ permalink raw reply related [flat|nested] only message in thread
only message in thread, other threads:[~2026-11-22 4:53 UTC | newest]
Thread overview: (only message) (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2026-11-22 4:53 [PATCH net-next v2] tg3: Propagate PHY reset retry failures in tg3_phy_reset_5703_4_5() Ian Rogers
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox