Date: Wed, 19 Dec 2007 10:21:41 -0800 From: Sean Bruno <sbruno@miralink.com> To: Hidetoshi Shimokawa <simokawa@FreeBSD.ORG> Cc: freebsd-firewire@freebsd.org Subject: Re: sbp_targ memory leak Message-ID: <47696135.8060508@miralink.com> In-Reply-To: <4768D2DD.2010500@miralink.com> References: <476610E5.2060108@miralink.com> <626eb4530712162258s4dfe1448o1102f20a623d3f95@mail.gmail.com> <476696C4.60408@miralink.com> <626eb4530712182320q237c344crd309893a82fe8ef8@mail.gmail.com> <4768D2DD.2010500@miralink.com>
next in thread | previous in thread | raw e-mail | index | archive | help
Sean Bruno wrote: > Hidetoshi Shimokawa wrote: >> I think you are right and page table is not freed when CAM_SEND_STATUS >> is not set. >> Maybe we should always free page tables if refcont == 0 rather than >> free in sbp_targ_send_status(). >> I thought about this some more this morning. What is causing us to allocate two page tables per XPT_CONT_TARGET_IO? The address printed out for the XPT_CONT_TARGET_IO call is identical, so I assume that these operations were for the same SCSI command. >> You patch is not just adding debug printfs, right? Yes, there is additions for XPT_SET_TRAN_SETTINGS /XPT_GET_TRAN_SETTINGS, additional CAM_SEND_SENSE flags and remove the requirement for target 0. Sean
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?47696135.8060508>