Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.kernel > #1217324

Re: [PATCH 15/16] mtd: mtdcore: fix initcall level

Show all headers | View raw

Am 01.09.2015 um 23:19 schrieb Brian Norris:
> Hi Alexander,
>
> No judgment here for the rest of this series, but for this patch:
>
> On Wed, Aug 26, 2015 at 02:28:27PM +0200, Alexander Holler wrote:
>> The mtd-core has to be initialized before other dependent mtd-drivers,
>> otherwise a crash might occur.
>>
>> Currently mtd_init() is called in the initcall-level device, which is the
>> same level where most mtd-drivers will end up. By luck this seemed to have
>> been called most of the time before other mtd-drivers without having been
>> explicitly enforced.
>
> I can't really speak for the original authors, but it does not appear to
> be entirely "by luck." Link order was one of the de facto ways to get
> this ordering (though it's not really a great one), and mtdcore was
> always linked first within the drivers/mtd/ directory structure.
>
> But that's just background, I think this is worth fixing anyway. It
> could, for instance, become a problem if drivers are located outside
> drivers/mtd/; I see random board files in arch/ that register with MTD,
> and I'm actually not sure how they have never tripped on this.

I've already found at least a half a dozen other drivers with the same 
problem through my shuffling of the drivers which all end up in the 
standard initcall level device. I'm aware that this is based on the link 
order, which itself is based on linker behaviour (and maybe other things 
like make or other build tools). I'm just calling it luck, because this 
is nowhere explicitly stated, at least I've never seen such a statement, 
neither in any source, nor somewhere in Documentation. So I've choosen 
the term "by luck" in order to provoke a bit (or to stimulate a 
discussion about that widespread problem).

>
>> But if mtd_init() is not called before a dependent
>> driver, a null-pointer exception might occur (e.g. because the mtd device
>> class isn't registered).
>>
>> To fix this, mtd-init() is moved to the initcall-level fs (right before
>> the standard initcall level device).
>
> Is there a good reason we shouldn't just make this a subsys_initcall()?
> That would match the naming better, and it mirrors what, e.g.,
> block/genhd uses. It would also allow flexibility if there are other
> current/future use cases that might need to sit between the subsystem
> and the drivers.

No real reason here. The names for the initcall levels seem to be 
outdated since a long time anyway, and I think just speaking about the 
numbers 1-7 (or 0-14) would be better anyways. The only reason why I've 
used the fs (sync) level is because I was too lazy to check out if 
mtdcore might depend on something else in any other level. Therefor I've 
used the one most close to were it was before.

Also I've an idea about how to fix that in general for all drivers (by 
using the same algorithm I've now introduced just for DT-based drivers 
with a device description). Wouldn't be that hard to use that for all 
drivers, but as I've written in a follow up to the introductory mail, 
one step after another.

Regards,

Alexander Holler
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

Back to linux.kernel | Previous | Next — Previous in thread | Next in thread | Find similar | Unroll thread

Thread

[PATCH 00/16] deps: deterministic driver initialization order Alexander Holler <holler@ahsoftware.de> - 2015-08-26 14:30 +0200
  [PATCH 01/16] deps: dtc: Automatically add new property 'dependencies' which contains a list of referenced phandles Alexander Holler <holler@ahsoftware.de> - 2015-08-26 14:30 +0200
  [PATCH 02/16] deps: dtc: Add option to print initialization order Alexander Holler <holler@ahsoftware.de> - 2015-08-26 14:30 +0200
  [PATCH 16/16] phy: phy-core: fix initcall level Alexander Holler <holler@ahsoftware.de> - 2015-08-26 14:40 +0200
  [PATCH 11/16] deps: WIP: generic: annotate some initcalls Alexander Holler <holler@ahsoftware.de> - 2015-08-26 14:40 +0200
  [PATCH 09/16] deps: dts: imx6q: make some remote-endpoints non-dependencies Alexander Holler <holler@ahsoftware.de> - 2015-08-26 14:40 +0200
  [PATCH 07/16] deps: add debug configuration options Alexander Holler <holler@ahsoftware.de> - 2015-08-26 14:40 +0200
  [PATCH 05/16] deps: introduce initcalls annotated with a struct device_driver Alexander Holler <holler@ahsoftware.de> - 2015-08-26 14:40 +0200
  [PATCH 04/16] deps: dtc: introduce new (virtual) property no-dependencies Alexander Holler <holler@ahsoftware.de> - 2015-08-26 14:40 +0200
  [PATCH 12/16] deps: WIP: imx: annotate some initcalls Alexander Holler <holler@ahsoftware.de> - 2015-08-26 14:40 +0200
  [PATCH 06/16] deps: deterministic driver initialization sequence based on dependencies from the DT Alexander Holler <holler@ahsoftware.de> - 2015-08-26 14:40 +0200
  [PATCH 08/16] deps: dts: kirkwood: dockstar: add dependency ehci -> usb power regulator Alexander Holler <holler@ahsoftware.de> - 2015-08-26 14:40 +0200
  [PATCH 13/16] deps: WIP: omap: annotate some initcalls Alexander Holler <holler@ahsoftware.de> - 2015-08-26 14:40 +0200
  [PATCH 03/16] deps: dtc: Add option to print dependency graph as dot (Graphviz) Alexander Holler <holler@ahsoftware.de> - 2015-08-26 14:40 +0200
  [PATCH 14/16] deps: WIP: kirkwood: annotate some initcalls Alexander Holler <holler@ahsoftware.de> - 2015-08-26 14:40 +0200
  [PATCH 10/16] deps: dts: omap: beagle: make some remote-endpoints non-dependencies Alexander Holler <holler@ahsoftware.de> - 2015-08-26 14:40 +0200
  [PATCH 15/16] mtd: mtdcore: fix initcall level Alexander Holler <holler@ahsoftware.de> - 2015-08-26 14:40 +0200
    Re: [PATCH 15/16] mtd: mtdcore: fix initcall level Brian Norris <computersforpeace@gmail.com> - 2015-09-01 23:20 +0200
      Re: [PATCH 15/16] mtd: mtdcore: fix initcall level Alexander Holler <holler@ahsoftware.de> - 2015-09-02 07:40 +0200
        Re: [PATCH 15/16] mtd: mtdcore: fix initcall level Alexander Holler <holler@ahsoftware.de> - 2015-09-04 06:10 +0200
  Re: [PATCH 00/16] deps: deterministic driver initialization order Alexander Holler <holler@ahsoftware.de> - 2015-08-27 21:10 +0200
    Re: [PATCH 00/16] deps: deterministic driver initialization order Alexander Holler <holler@ahsoftware.de> - 2015-08-27 21:20 +0200
  deps: update in regard to parallel initialization of static linked  drivers Alexander Holler <holler@ahsoftware.de> - 2015-09-04 11:20 +0200

csiph-web