[src/jdolecek-ncq]: src/sys/dev/ata more notes, esp. PMP on siisata doesn't w...

[jdolecek <jdolecek%NetBSD.org@localhost>]

details:   https://anonhg.NetBSD.org/src/rev/8bf3d9e239c6
branches:  jdolecek-ncq
changeset: 352675:8bf3d9e239c6
user:      jdolecek <jdolecek%NetBSD.org@localhost>
date:      Sat Jun 17 19:24:26 2017 +0000

description:
more notes, esp. PMP on siisata doesn't work + more problems with downsize

diffstat:

 sys/dev/ata/TODO.ncq |  11 ++++++++---
 1 files changed, 8 insertions(+), 3 deletions(-)

diffs (32 lines):

diff -r d0dce1b72d79 -r 8bf3d9e239c6 sys/dev/ata/TODO.ncq
--- a/sys/dev/ata/TODO.ncq      Sat Jun 17 14:01:36 2017 +0000
+++ b/sys/dev/ata/TODO.ncq      Sat Jun 17 19:24:26 2017 +0000
@@ -1,13 +1,15 @@
 Bugs
 ----
 
-test crashdump with ahci, pciide, mvsata
-- jdolecek: broken for various reasons, working on fix
+fix crashdump for mvsata - request times out (maybe same as HEAD?)
 
-siisata - fix all new XXX and unmergable bits
+siisata - fix all new XXX and unmergable bits, fix PMP
 
 test crashdump with siisata
 
+reset channel and PMP doesn't work with siisata - siisata_reset_drive()
+uses ata_get_xfer() and this can clash on slot 0 with on-stack xfers
+
 test wd* at umass?, confirm the ata_channel kludge works
 
 is ata_exec_xfer() + POLL safe wrt. more outstanding I/Os? why is it waiting
@@ -24,6 +26,9 @@
 drive rescaned after detach of non-NCQ drive
 - careful with PMP, must be minimum of openings supported by drives
   attached to the same channel
+- the downsize can leak transfers with PMP if other device happens
+  to have active or pending transfers (e.g. when non-NCQ device is attached
+  while there is already NCQ device present)
 
 Other random notes (do outside the NCQ branch):
 -----------------------------------------------------