netdev - Re: [PATCH v3 6/7] venus: Make debug infrastructure more flexible

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Thu, 11 Jun 2020 13:18:17 +0100
From:   Daniel Thompson <daniel.thompson@...aro.org>
To:     Stanimir Varbanov <stanimir.varbanov@...aro.org>
Cc:     Joe Perches <joe@...ches.com>,
        Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
        linux-doc@...r.kernel.org, linux-kernel@...r.kernel.org,
        linux-media@...r.kernel.org, linux-arm-msm@...r.kernel.org,
        linux-btrfs@...r.kernel.org, linux-acpi@...r.kernel.org,
        netdev@...r.kernel.org, Jason Baron <jbaron@...mai.com>
Subject: Re: [PATCH v3 6/7] venus: Make debug infrastructure more flexible

On Thu, Jun 11, 2020 at 02:31:07PM +0300, Stanimir Varbanov wrote:
> On 6/11/20 1:52 PM, Daniel Thompson wrote:
> > On Wed, Jun 10, 2020 at 11:42:43PM -0700, Joe Perches wrote:
> >> On Thu, 2020-06-11 at 08:26 +0200, Greg Kroah-Hartman wrote:
> >>> On Wed, Jun 10, 2020 at 01:23:56PM -0700, Joe Perches wrote:
> >>>> On Wed, 2020-06-10 at 12:49 -0700, Joe Perches wrote:
> >>>>> On Wed, 2020-06-10 at 15:37 +0200, Greg Kroah-Hartman wrote:
> >>>>>> Please work with the infrastructure we have, we have spent a lot of time
> >>>>>> and effort to make it uniform to make it easier for users and
> >>>>>> developers.
> >>>>>
> >>>>> Not quite.
> >>>>>
> >>>>> This lack of debug grouping by type has been a
> >>>>> _long_ standing issue with drivers.
> >>>>>
> >>>>>> Don't regress and try to make driver-specific ways of doing
> >>>>>> things, that way lies madness...
> >>>>>
> >>>>> It's not driver specific, it allows driver developers to
> >>>>> better isolate various debug states instead of keeping
> >>>>> lists of specific debug messages and enabling them
> >>>>> individually.
> >>>>
> >>>> For instance, look at the homebrew content in
> >>>> drivers/gpu/drm/drm_print.c that does _not_ use
> >>>> dynamic_debug.
> >>>>
> >>>> MODULE_PARM_DESC(debug, "Enable debug output, where each bit enables a debug category.\n"
> >>>> "\t\tBit 0 (0x01)  will enable CORE messages (drm core code)\n"
> >>>> "\t\tBit 1 (0x02)  will enable DRIVER messages (drm controller code)\n"
> >>>> "\t\tBit 2 (0x04)  will enable KMS messages (modesetting code)\n"
> >>>> "\t\tBit 3 (0x08)  will enable PRIME messages (prime code)\n"
> >>>> "\t\tBit 4 (0x10)  will enable ATOMIC messages (atomic code)\n"
> >>>> "\t\tBit 5 (0x20)  will enable VBL messages (vblank code)\n"
> >>>> "\t\tBit 7 (0x80)  will enable LEASE messages (leasing code)\n"
> >>>> "\t\tBit 8 (0x100) will enable DP messages (displayport code)");
> >>>> module_param_named(debug, __drm_debug, int, 0600);
> >>>>
> >>>> void drm_dev_dbg(const struct device *dev, enum drm_debug_category category,
> >>>> 		 const char *format, ...)
> >>>> {
> >>>> 	struct va_format vaf;
> >>>> 	va_list args;
> >>>>
> >>>> 	if (!drm_debug_enabled(category))
> >>>> 		return;
> >>>
> >>> Ok, and will this proposal be able to handle stuff like this?
> >>
> >> Yes, that's the entire point.
> > 
> > Currently I think there not enough "levels" to map something like
> > drm.debug to the new dyn dbg feature. I don't think it is intrinsic
> > but I couldn't find the bit of the code where the 5-bit level in struct
> > _ddebug is converted from a mask to a bit number and vice-versa.
> 
> Here [1] is Joe's initial suggestion. But I decided that bitmask is a
> good start for the discussion.
> 
> I guess we can add new member uint "level" in struct _ddebug so that we
> can cover more "levels" (types, groups).

I don't think it is allocating only 5 bits that is the problem!

The problem is that those 5 bits need not be encoded as a bitmask by
dyndbg, that can simply be the category code for the message. They only
need be converted into a mask when we compare them to the mask provided
by the user.


Daniel.