ata: libata-eh: do not thaw the port twice in ata_eh_reset()
authorNiklas Cassel <niklas.cassel@wdc.com>
Wed, 13 Sep 2023 22:19:17 +0000 (00:19 +0200)
committerDamien Le Moal <dlemoal@kernel.org>
Sat, 16 Sep 2023 12:11:28 +0000 (21:11 +0900)
commit 1e641060c4b5 ("libata: clear eh_info on reset completion") added
a workaround that broke the retry mechanism in ATA EH.

Tejun himself suggested to remove this workaround when it was identified
to cause additional problems:
https://lore.kernel.org/linux-ide/20110426135027.GI878@htj.dyndns.org/

He even said:
"Hmm... it seems I wasn't thinking straight when I added that work around."
https://lore.kernel.org/linux-ide/20110426155229.GM878@htj.dyndns.org/

While removing the workaround solved the issue, however, the workaround was
kept to avoid "spurious hotplug events during reset", and instead another
workaround was added on top of the existing workaround in commit
8c56cacc724c ("libata: fix unexpectedly frozen port after ata_eh_reset()").

Because these IRQs happened when the port was frozen, we know that they
were actually a side effect of PxIS and IS.IPS(x) not being cleared before
the COMRESET. This is now done in commit 94152042eaa9 ("ata: libahci: clear
pending interrupt status"), so these workarounds can now be removed.

Since commit 1e641060c4b5 ("libata: clear eh_info on reset completion") has
now been reverted, the ATA EH retry mechanism is functional again, so there
is once again no need to thaw the port more than once in ata_eh_reset().

This reverts "the workaround on top of the workaround" introduced in commit
8c56cacc724c ("libata: fix unexpectedly frozen port after ata_eh_reset()").

Signed-off-by: Niklas Cassel <niklas.cassel@wdc.com>
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
drivers/ata/libata-eh.c

index 5c493b6316ebf5bd32af91d6a9300eaf71ec82bb..4cf4f57e57b8bdf06b5353e01ea4236c45f705e4 100644 (file)
@@ -2803,9 +2803,6 @@ int ata_eh_reset(struct ata_link *link, int classify,
                slave->eh_info.serror = 0;
        spin_unlock_irqrestore(link->ap->lock, flags);
 
-       if (ata_port_is_frozen(ap))
-               ata_eh_thaw_port(ap);
-
        /*
         * Make sure onlineness and classification result correspond.
         * Hotplug could have happened during reset and some