Re: [PATCH v9 08/12] device core: Introduce DMA range map, supplanting dma_pfn_offset

From: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
To: Jim Quinlan <james.quinlan@broadcom.com>,
	Rob Herring <robh+dt@kernel.org>
Cc: Rich Felker <dalias@libc.org>,
	"open list:SUPERH" <linux-sh@vger.kernel.org>,
	David Airlie <airlied@linux.ie>, PCI <linux-pci@vger.kernel.org>,
	Hanjun Guo <guohanjun@huawei.com>,
	"open list:REMOTE PROCESSOR \(REMOTEPROC\) SUBSYSTEM"
	<linux-remoteproc@vger.kernel.org>,
	Andy Shevchenko <andriy.shevchenko@linux.intel.com>,
	Julien Grall <julien.grall@arm.com>,
	Heikki Krogerus <heikki.krogerus@linux.intel.com>,
	"H. Peter Anvin" <hpa@zytor.com>, Will Deacon <will@kernel.org>,
	Christoph Hellwig <hch@lst.de>,
	"open list:STAGING SUBSYSTEM" <devel@driverdev.osuosl.org>,
	Jean-Philippe Brucker <jean-philippe@linaro.org>,
	Yoshinori Sato <ysato@users.sourceforge.jp>,
	Frank Rowand <frowand.list@gmail.com>,
	"maintainer:X86 ARCHITECTURE \(32-BIT AND 64-BIT\)"
	<x86@kernel.org>, Russell King <linux@armlinux.org.uk>,
	"open list:ACPI FOR ARM64 \(ACPI/arm64\)"
	<linux-acpi@vger.kernel.org>, Chen-Yu Tsai <wens@csie.org>,
	Ingo Molnar <mingo@redhat.com>,
	"maintainer:BROADCOM BCM7XXX ARM ARCHITECTURE"
	<bcm-kernel-feedback-list@broadcom.com>,
	Alan Stern <stern@rowland.harvard.edu>,
	Len Brown <lenb@kernel.org>, Ohad Ben-Cohen <ohad@wizery.com>,
	"open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE"
	<devicetree@vger.kernel.org>,
	Philipp Zabel <p.zabel@pengutronix.de>,
	Arnd Bergmann <arnd@arndb.de>,
	Suzuki K Poulose <suzuki.poulose@arm.com>,
	Maxime Ripard <mripard@kernel.org>,
	Florian Fainelli <f.fainelli@gmail.com>,
	Borislav Petkov <bp@alien8.de>,
	"open list:DRM DRIVERS FOR ALLWINNER A10"
	<dri-devel@lists.freedesktop.org>,
	Yong Deng <yong.deng@magewell.com>,
	Santosh Shilimkar <ssantosh@kernel.org>,
	Bjorn Helgaas <bhelgaas@google.com>,
	Thomas Gleixner <tglx@linutronix.de>,
	Mauro Carvalho Chehab <mchehab@kernel.org>,
	"moderated list:ARM PORT" <linux-arm-kernel@lists.infradead.org>,
	Saravana Kannan <saravanak@google.com>,
	Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
	Oliver Neukum <oneukum@suse.com>,
	"Rafael J. Wysocki" <rjw@rjwysocki.net>,
	open list <linux-kernel@vger.kernel.org>,
	Paul Kocialkowski <paul.kocialkowski@bootlin.com>,
	"open list:IOMMU DRIVERS" <iommu@lists.linux-foundation.org>,
	"open list:USB SUBSYSTEM" <linux-usb@vger.kernel.org>,
	Stefano Stabellini <sstabellini@kernel.org>,
	Daniel Vetter <daniel@ffwll.ch>,
	Sudeep Holla <sudeep.holla@arm.com>,
	"open list:ALLWINNER A10 CSI DRIVER"
	<linux-media@vger.kernel.org>,
	Robin Murphy <robin.murphy@arm.com>,
	"moderated list:BROADCOM BCM2711/BCM2835 ARM ARCHITECTURE"
	<linux-rpi-kernel@lists.infradead.org>
Subject: Re: [PATCH v9 08/12] device core: Introduce DMA range map, supplanting dma_pfn_offset
Date: Thu, 30 Jul 2020 19:02:37 +0200	[thread overview]
Message-ID: <1f3c4641a3058159a2923c1f5833a10b413747bf.camel@suse.de> (raw)
In-Reply-To: <CA+-6iNyi4c+ZUZhe=Zn5qX5GiGQVLR+BOcXmXSJkooX=S9=C0A@mail.gmail.com>

On Thu, 2020-07-30 at 12:44 -0400, Jim Quinlan wrote:
> On Wed, Jul 29, 2020 at 10:28 AM Rob Herring <robh+dt@kernel.org> wrote:
> > On Wed, Jul 29, 2020 at 12:19 AM Christoph Hellwig <hch@lst.de> wrote:
> > > On Tue, Jul 28, 2020 at 02:24:51PM -0400, Jim Quinlan wrote:
> > > > I started using devm_kcalloc() but at least two reviewers convinced me
> > > > to just use kcalloc().  In addition, when I was using devm_kcalloc()
> > > > it was awkward because 'dev' is not available to this function.
> > > > 
> > > > It comes down to whether unbind/binding the device N times is actually
> > > > a reasonable usage.  As for my experience I've seen two cases: (1) my
> > > > overnight "bind/unbind the PCIe RC driver" script, and we have a
> > > > customer who does an unbind/bind as a hail mary to bring back life to
> > > > their dead EP device.  If the latter case happens repeatedly, there
> > > > are bigger problems.
> > > 
> > > We can't just leak the allocations.  Do you have a pointer to the
> > > arguments against managed resources?  I'm generally not a huge fan
> > > of the managed resources, but for a case like this they actually seem
> > > useful.  If we don't use the managed resources we'll at leat need
> > > to explicitly free the resources when freeing the device.
> > 
> > The lifetime for devm_kcalloc may not be what we want here. devm
> > allocs are freed on probe fail or remove, not on freeing the device
> > (there is a just in case free there too though).
> 
> What do you suggest doing as an alternative?

I haven't given the idea much thought, but how about using a kref in the first
element of your bus_dma_region array. It should be increased upon assigning it
to a device and decreased it upon destroying it (triggering the free once it
hits 0). It would take care of this memory leak, but also useful where you're
sharing bus_dma_regions between devices.

Regards,
Nicolas

_______________________________________________
iommu mailing list
iommu@lists.linux-foundation.org
https://lists.linuxfoundation.org/mailman/listinfo/iommu