[PATCH net v3] net: macb: fix ordering around PTP timestamp read

PTP_SYS_OFFSET_EXTENDED returns system timestamps that do not correctly
bracket the PHC register read on MACB/GEM. On a Raspberry Pi 5, the
returned interval can be as short as 37 ns, while an ordered register
read takes approximately 1 us. This biases the midpoint used by phc2sys,
causing CLOCK_REALTIME to run approximately 0.5 us ahead when synchronized
to the PHC.

gem_tsu_get_time() reads the nanoseconds register using the driver's
relaxed MMIO accessor. On weakly ordered systems, the subsequent system
timestamp can be taken before the register read completes. The internal
smp_rmb() in the pre-timestamp path also does not guarantee ordering
against the subsequent MMIO read.

Add rmb() before and after the bracketed nanoseconds read in both the
normal and seconds rollover paths so the system timestamps bracket the
PHC read. Adding the post-read barrier increases the minimum interval on
the same Raspberry Pi 5 to approximately 1 us.

Fixes: e51bb5c2784c ("net: macb: ptp: Switch to gettimex64() interface")
Tested-by: Nicolai Buchwitz <redacted> # Raspberry Pi CM5, min bracket 37 ns -> 981 ns
Reviewed-by: Nicolai Buchwitz <redacted>
Reviewed-by: Théo Lebrun <theo.lebrun@bootlin.com>
Assisted-by: LLM
Signed-off-by: James Clark <redacted>
---
Changes in v3:
- Guard the barriers with if (sts), following Jakub's feedback.
- Reduce the comments following Théo's feedback, and focus on relevant
  point following Jakub's feedback.
- Add Théo Lebrun's Reviewed-by from v2.
- Add Assisted-by: LLM.

v2: https://lore.kernel.org/netdev/20260910040658.18359-1-jjc@jclark.com/ (local)
v1: https://lore.kernel.org/netdev/20260908053150.28694-1-jjc@jclark.com/ (local)

Nicolai's testing and review covered v1; Théo's review covered v2.

 drivers/net/ethernet/cadence/macb_ptp.c | 9 +++++++++
 1 file changed, 9 insertions(+)

diff --git a/drivers/net/ethernet/cadence/macb_ptp.c b/drivers/net/ethernet/cadence/macb_ptp.c
index e5195d7da..a9ac09900 100644
--- a/drivers/net/ethernet/cadence/macb_ptp.c
+++ b/drivers/net/ethernet/cadence/macb_ptp.c

@@ -50,7 +50,12 @@ static int gem_tsu_get_time(struct ptp_clock_info *ptp, struct timespec64 *ts,
 
 	spin_lock_irqsave(&bp->tsu_clk_lock, flags);
 	ptp_read_system_prets(sts);
+	/* explicit barriers are needed because gem_readl() is relaxed */
+	if (sts)
+		rmb();
 	first = gem_readl(bp, TN);
+	if (sts)
+		rmb();
 	ptp_read_system_postts(sts);
 	secl = gem_readl(bp, TSL);
 	sech = gem_readl(bp, TSH);

@@ -62,7 +67,11 @@ static int gem_tsu_get_time(struct ptp_clock_info *ptp, struct timespec64 *ts,
 		 * (assume all done within 1s)
 		 */
 		ptp_read_system_prets(sts);
+		if (sts)
+			rmb();
 		ts->tv_nsec = gem_readl(bp, TN);
+		if (sts)
+			rmb();
 		ptp_read_system_postts(sts);
 		secl = gem_readl(bp, TSL);
 		sech = gem_readl(bp, TSH);