Re: port-arm/54199: arm bus_dmamap_sync assertion failure

[Izumi Tsutsui <tsutsui%ceres.dti.ne.jp@localhost>]

The following reply was made to PR port-arm/54199; it has been noted by GNATS.

From: Izumi Tsutsui <tsutsui%ceres.dti.ne.jp@localhost>
To: gnats-bugs%netbsd.org@localhost
Cc: thorpej%me.com@localhost, tsutsui%ceres.dti.ne.jp@localhost
Subject: Re: port-arm/54199: arm bus_dmamap_sync assertion failure
Date: Tue, 14 May 2019 00:02:14 +0900

 > System was copying a large amount of data (comprised mostly of small files) from one USB-attached SSD to another.  "tar" command output was being relayed over an ssh session.  At some point, the kernel panic'd:
 > 
 > [ 1735156.504667] panic: kernel diagnostic assertion "offset < map->dm_mapsize" failed: file "../../../../arch/arm/arm32/bus_dma.c", line 1051 offset 0 mapsize 0
 
 I have a question:
  Is it a valid operation to call bus_dmamap_load(9) and bus_dmamap_sync(9)
  with dmamap dm_mapsize == 0?
 
 In that case, the assertion "offset < map->dm_mapsize" always fails
 even in case of offset==0.
 
 Back in 2011 there was a report that src/sys/dev/ic/ahcisata_core.c
 could cause zero-sized dmamap:
  http://www.re.soum.co.jp/~fukaumi/ml/netbsd/201101/msg00003.html
 and I committed a fix to avoid such calls.
  http://mail-index.netbsd.org/source-changes/2011/01/10/msg017063.html
 
 Maybe some other drivers have the similar ops, but if it's vaild ops
 all assertion should be fixed.
 
 ---
 Izumi Tsutsui