intel-gfx.lists.freedesktop.org archive mirror
 help / color / mirror / Atom feed
From: Jason Ekstrand <jason@jlekstrand.net>
To: Tvrtko Ursulin <tvrtko.ursulin@linux.intel.com>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>,
	Intel Graphics Development <intel-gfx@lists.freedesktop.org>,
	DRI Development <dri-devel@lists.freedesktop.org>,
	Daniel Vetter <daniel.vetter@intel.com>
Subject: Re: [Intel-gfx] [PATCH 04/10] drm/i915: move intel_context slab to direct module init/exit
Date: Mon, 26 Jul 2021 11:20:28 -0500	[thread overview]
Message-ID: <CAOFGe95G3bYWxZ_EwPEFzxf10vf1T3Zdyez_fbh=imcDj+TSHw@mail.gmail.com> (raw)
In-Reply-To: <031997ab-5568-9dbd-fcee-b4f820a32632@linux.intel.com>

On Mon, Jul 26, 2021 at 11:08 AM Tvrtko Ursulin
<tvrtko.ursulin@linux.intel.com> wrote:
> On 26/07/2021 16:42, Jason Ekstrand wrote:
> > On Mon, Jul 26, 2021 at 10:30 AM Jason Ekstrand <jason@jlekstrand.net> wrote:
> >>
> >> On Mon, Jul 26, 2021 at 3:35 AM Tvrtko Ursulin
> >> <tvrtko.ursulin@linux.intel.com> wrote:
> >>>
> >>>
> >>> On 23/07/2021 20:29, Daniel Vetter wrote:
> >>>> With the global kmem_cache shrink infrastructure gone there's nothing
> >>>> special and we can convert them over.
> >>>>
> >>>> I'm doing this split up into each patch because there's quite a bit of
> >>>> noise with removing the static global.slab_ce to just a
> >>>> slab_ce.
> >>>>
> >>>> Cc: Jason Ekstrand <jason@jlekstrand.net>
> >>>> Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
> >>>> ---
> >>>>    drivers/gpu/drm/i915/gt/intel_context.c | 25 ++++++++-----------------
> >>>>    drivers/gpu/drm/i915/gt/intel_context.h |  3 +++
> >>>>    drivers/gpu/drm/i915/i915_globals.c     |  2 --
> >>>>    drivers/gpu/drm/i915/i915_globals.h     |  1 -
> >>>>    drivers/gpu/drm/i915/i915_pci.c         |  2 ++
> >>>>    5 files changed, 13 insertions(+), 20 deletions(-)
> >>>>
> >>>> diff --git a/drivers/gpu/drm/i915/gt/intel_context.c b/drivers/gpu/drm/i915/gt/intel_context.c
> >>>> index baa05fddd690..283382549a6f 100644
> >>>> --- a/drivers/gpu/drm/i915/gt/intel_context.c
> >>>> +++ b/drivers/gpu/drm/i915/gt/intel_context.c
> >>>> @@ -7,7 +7,6 @@
> >>>>    #include "gem/i915_gem_pm.h"
> >>>>
> >>>>    #include "i915_drv.h"
> >>>> -#include "i915_globals.h"
> >>>>    #include "i915_trace.h"
> >>>>
> >>>>    #include "intel_context.h"
> >>>> @@ -15,14 +14,11 @@
> >>>>    #include "intel_engine_pm.h"
> >>>>    #include "intel_ring.h"
> >>>>
> >>>> -static struct i915_global_context {
> >>>> -     struct i915_global base;
> >>>> -     struct kmem_cache *slab_ce;
> >>>> -} global;
> >>>> +struct kmem_cache *slab_ce;
> >>
> >> Static?  With that,
> >>
> >> Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
> >>
> >>>>
> >>>>    static struct intel_context *intel_context_alloc(void)
> >>>>    {
> >>>> -     return kmem_cache_zalloc(global.slab_ce, GFP_KERNEL);
> >>>> +     return kmem_cache_zalloc(slab_ce, GFP_KERNEL);
> >>>>    }
> >>>>
> >>>>    static void rcu_context_free(struct rcu_head *rcu)
> >>>> @@ -30,7 +26,7 @@ static void rcu_context_free(struct rcu_head *rcu)
> >>>>        struct intel_context *ce = container_of(rcu, typeof(*ce), rcu);
> >>>>
> >>>>        trace_intel_context_free(ce);
> >>>> -     kmem_cache_free(global.slab_ce, ce);
> >>>> +     kmem_cache_free(slab_ce, ce);
> >>>>    }
> >>>>
> >>>>    void intel_context_free(struct intel_context *ce)
> >>>> @@ -410,22 +406,17 @@ void intel_context_fini(struct intel_context *ce)
> >>>>        i915_active_fini(&ce->active);
> >>>>    }
> >>>>
> >>>> -static void i915_global_context_exit(void)
> >>>> +void i915_context_module_exit(void)
> >>>>    {
> >>>> -     kmem_cache_destroy(global.slab_ce);
> >>>> +     kmem_cache_destroy(slab_ce);
> >>>>    }
> >>>>
> >>>> -static struct i915_global_context global = { {
> >>>> -     .exit = i915_global_context_exit,
> >>>> -} };
> >>>> -
> >>>> -int __init i915_global_context_init(void)
> >>>> +int __init i915_context_module_init(void)
> >>>>    {
> >>>> -     global.slab_ce = KMEM_CACHE(intel_context, SLAB_HWCACHE_ALIGN);
> >>>> -     if (!global.slab_ce)
> >>>> +     slab_ce = KMEM_CACHE(intel_context, SLAB_HWCACHE_ALIGN);
> >>>> +     if (!slab_ce)
> >>>>                return -ENOMEM;
> >>>>
> >>>> -     i915_global_register(&global.base);
> >>>>        return 0;
> >>>>    }
> >>>>
> >>>> diff --git a/drivers/gpu/drm/i915/gt/intel_context.h b/drivers/gpu/drm/i915/gt/intel_context.h
> >>>> index 974ef85320c2..a0ca82e3c40d 100644
> >>>> --- a/drivers/gpu/drm/i915/gt/intel_context.h
> >>>> +++ b/drivers/gpu/drm/i915/gt/intel_context.h
> >>>> @@ -30,6 +30,9 @@ void intel_context_init(struct intel_context *ce,
> >>>>                        struct intel_engine_cs *engine);
> >>>>    void intel_context_fini(struct intel_context *ce);
> >>>>
> >>>> +void i915_context_module_exit(void);
> >>>> +int i915_context_module_init(void);
> >>>> +
> >>>>    struct intel_context *
> >>>>    intel_context_create(struct intel_engine_cs *engine);
> >>>>
> >>>> diff --git a/drivers/gpu/drm/i915/i915_globals.c b/drivers/gpu/drm/i915/i915_globals.c
> >>>> index 3de7cf22ec76..d36eb7dc40aa 100644
> >>>> --- a/drivers/gpu/drm/i915/i915_globals.c
> >>>> +++ b/drivers/gpu/drm/i915/i915_globals.c
> >>>> @@ -7,7 +7,6 @@
> >>>>    #include <linux/slab.h>
> >>>>    #include <linux/workqueue.h>
> >>>>
> >>>> -#include "gem/i915_gem_context.h"
> >>>>    #include "gem/i915_gem_object.h"
> >>>>    #include "i915_globals.h"
> >>>>    #include "i915_request.h"
> >>>> @@ -32,7 +31,6 @@ static void __i915_globals_cleanup(void)
> >>>>    }
> >>>>
> >>>>    static __initconst int (* const initfn[])(void) = {
> >>>> -     i915_global_context_init,
> >>>>        i915_global_gem_context_init,
> >>>>        i915_global_objects_init,
> >>>>        i915_global_request_init,
> >>>> diff --git a/drivers/gpu/drm/i915/i915_globals.h b/drivers/gpu/drm/i915/i915_globals.h
> >>>> index d80901ba75e3..60daa738a188 100644
> >>>> --- a/drivers/gpu/drm/i915/i915_globals.h
> >>>> +++ b/drivers/gpu/drm/i915/i915_globals.h
> >>>> @@ -23,7 +23,6 @@ int i915_globals_init(void);
> >>>>    void i915_globals_exit(void);
> >>>>
> >>>>    /* constructors */
> >>>> -int i915_global_context_init(void);
> >>>>    int i915_global_gem_context_init(void);
> >>>>    int i915_global_objects_init(void);
> >>>>    int i915_global_request_init(void);
> >>>> diff --git a/drivers/gpu/drm/i915/i915_pci.c b/drivers/gpu/drm/i915/i915_pci.c
> >>>> index f9527269e30a..266618157775 100644
> >>>> --- a/drivers/gpu/drm/i915/i915_pci.c
> >>>> +++ b/drivers/gpu/drm/i915/i915_pci.c
> >>>> @@ -33,6 +33,7 @@
> >>>>    #include "i915_active.h"
> >>>>    #include "i915_buddy.h"
> >>>>    #include "i915_drv.h"
> >>>> +#include "gem/i915_gem_context.h"
> >>>
> >>> It's a bit ugly to go to a design where i915_pci.c has to include so
> >>> many random parts of i915. IMO for a complex driver like i915,
> >>> compartmentalizing so much knowledge about the internals was better
> >>> inside the globals layer.
> >>
> >> I agree that i915_pci feels like the wrong place to put this but I
> >> don't think that's so much because globals don't belong in i915_pci
> >> but because i915_init/exit don't belong there.  Maybe, once this is
> >> all said and done (or at the start of the series), we should move
> >> i915_init/exit to i915_drv.c?  Of course, there's a bunch of PCI
> >> probing stuff in i915_drv.c so..... yeah.... our organization is
> >> pretty busted.
> >
> > To put a finer point on this, the new "design" is really to have a
> > single flat list instead of two, one nested inside the other.  There's
> > nothing wrong with that at all.  The fact that all this stuff now
> > lives in i915_pci.c is ugly.  But, as I said, that's kind-of an
> > accident of history because that's where i915_init() and i915_exit()
> > currently live.  We should just move the lot to i915_drv.c.
>
> Hmm.. on one hand it does sounds better to move to i915_drv.c, but is it
> just because all these new include directive are so visibly out of place
> in i915_pci.c?
>
> Perhaps we need i915_module.c and then i915_globals is a completely fine
> concept. Desired IMO even since we have to avoid globals in general
> (multi-gpu) so it sticks out nicely that all that is allowed to be
> global has a special place.
>
> And i915_drv.c can remain being about a driver instance as bound to one GPU.

Is i915_drv.c about a single instance bound to a single GPU?  If so,
then, yeah, maybe not the right place.  Maybe a i915_module.c would be
better.  It's all different shades of shed paint.

--Jason

> That feels like the best of both worlds to me.
>
> Regards,
>
> Tvrtko
>
> >
> >> --Jason
> >>
> >>> Maybe add a cover letter to explain the perceived pros and cons and
> >>> thinking in general?
> >>>
> >>> Regards,
> >>>
> >>> Tvrtko
> >>>
> >>>>    #include "i915_perf.h"
> >>>>    #include "i915_globals.h"
> >>>>    #include "i915_selftest.h"
> >>>> @@ -1297,6 +1298,7 @@ static const struct {
> >>>>        { i915_check_nomodeset, NULL },
> >>>>        { i915_active_module_init, i915_active_module_exit },
> >>>>        { i915_buddy_module_init, i915_buddy_module_exit },
> >>>> +     { i915_context_module_init, i915_context_module_exit },
> >>>>        { i915_globals_init, i915_globals_exit },
> >>>>        { i915_mock_selftests, NULL },
> >>>>        { i915_pmu_init, i915_pmu_exit },
> >>>>
> >>> _______________________________________________
> >>> Intel-gfx mailing list
> >>> Intel-gfx@lists.freedesktop.org
> >>> https://lists.freedesktop.org/mailman/listinfo/intel-gfx
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

  reply	other threads:[~2021-07-26 16:20 UTC|newest]

Thread overview: 46+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-07-23 19:29 [Intel-gfx] [PATCH 01/10] drm/i915: Check for nomodeset in i915_init() first Daniel Vetter
2021-07-23 19:29 ` [Intel-gfx] [PATCH 02/10] drm/i915: move i915_active slab to direct module init/exit Daniel Vetter
2021-07-26 15:24   ` Jason Ekstrand
2021-07-23 19:29 ` [Intel-gfx] [PATCH 03/10] drm/i915: move i915_buddy " Daniel Vetter
2021-07-24  7:38   ` kernel test robot
2021-07-24  7:38   ` [Intel-gfx] [RFC PATCH] drm/i915: slab_blocks can be static kernel test robot
2021-07-26 15:26   ` [Intel-gfx] [PATCH 03/10] drm/i915: move i915_buddy slab to direct module init/exit Jason Ekstrand
2021-07-23 19:29 ` [Intel-gfx] [PATCH 04/10] drm/i915: move intel_context " Daniel Vetter
2021-07-24 11:45   ` kernel test robot
2021-07-24 11:45   ` [Intel-gfx] [RFC PATCH] drm/i915: slab_ce can be static kernel test robot
2021-07-26  8:35   ` [Intel-gfx] [PATCH 04/10] drm/i915: move intel_context slab to direct module init/exit Tvrtko Ursulin
2021-07-26 15:30     ` Jason Ekstrand
2021-07-26 15:42       ` Jason Ekstrand
2021-07-26 16:08         ` Tvrtko Ursulin
2021-07-26 16:20           ` Jason Ekstrand [this message]
2021-07-26 16:31             ` Tvrtko Ursulin
2021-07-26 18:17               ` Jason Ekstrand
2021-07-27 10:14                 ` Tvrtko Ursulin
2021-07-23 19:29 ` [Intel-gfx] [PATCH 05/10] drm/i915: move gem_context " Daniel Vetter
2021-07-24 14:50   ` kernel test robot
2021-07-24 14:50   ` [Intel-gfx] [RFC PATCH] drm/i915: slab_luts can be static kernel test robot
2021-07-26 15:35   ` [Intel-gfx] [PATCH 05/10] drm/i915: move gem_context slab to direct module init/exit Jason Ekstrand
2021-07-23 19:29 ` [Intel-gfx] [PATCH 06/10] drm/i915: move gem_objects " Daniel Vetter
2021-07-24 18:23   ` kernel test robot
2021-07-24 18:23   ` [Intel-gfx] [RFC PATCH] drm/i915: slab_objects can be static kernel test robot
2021-07-26 15:39   ` [Intel-gfx] [PATCH 06/10] drm/i915: move gem_objects slab to direct module init/exit Jason Ekstrand
2021-07-23 19:29 ` [Intel-gfx] [PATCH 07/10] drm/i915: move request slabs " Daniel Vetter
2021-07-24 21:58   ` kernel test robot
2021-07-24 21:58   ` [Intel-gfx] [RFC PATCH] drm/i915: slab_requests can be static kernel test robot
2021-07-26 15:46   ` [Intel-gfx] [PATCH 07/10] drm/i915: move request slabs to direct module init/exit Jason Ekstrand
2021-07-23 19:29 ` [Intel-gfx] [PATCH 08/10] drm/i915: move scheduler " Daniel Vetter
2021-07-25  1:23   ` kernel test robot
2021-07-25  1:23   ` [Intel-gfx] [RFC PATCH] drm/i915: slab_dependencies can be static kernel test robot
2021-07-26 15:47   ` [Intel-gfx] [PATCH 08/10] drm/i915: move scheduler slabs to direct module init/exit Jason Ekstrand
2021-07-23 19:29 ` [Intel-gfx] [PATCH 09/10] drm/i915: move vma slab " Daniel Vetter
2021-07-25  4:04   ` kernel test robot
2021-07-25  4:04   ` [Intel-gfx] [RFC PATCH] drm/i915: slab_vmas can be static kernel test robot
2021-07-26 15:50   ` [Intel-gfx] [PATCH 09/10] drm/i915: move vma slab to direct module init/exit Jason Ekstrand
2021-07-23 19:29 ` [Intel-gfx] [PATCH 10/10] drm/i915: Remove i915_globals Daniel Vetter
2021-07-26 15:51   ` Jason Ekstrand
2021-07-27 11:34     ` Daniel Vetter
2021-07-23 21:55 ` [Intel-gfx] ✗ Fi.CI.CHECKPATCH: warning for series starting with [01/10] drm/i915: Check for nomodeset in i915_init() first Patchwork
2021-07-23 21:56 ` [Intel-gfx] ✗ Fi.CI.SPARSE: " Patchwork
2021-07-23 22:29 ` [Intel-gfx] ✓ Fi.CI.BAT: success " Patchwork
2021-07-24  9:00 ` [Intel-gfx] ✓ Fi.CI.IGT: " Patchwork
2021-07-26 15:23 ` [Intel-gfx] [PATCH 01/10] " Jason Ekstrand

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to='CAOFGe95G3bYWxZ_EwPEFzxf10vf1T3Zdyez_fbh=imcDj+TSHw@mail.gmail.com' \
    --to=jason@jlekstrand.net \
    --cc=daniel.vetter@ffwll.ch \
    --cc=daniel.vetter@intel.com \
    --cc=dri-devel@lists.freedesktop.org \
    --cc=intel-gfx@lists.freedesktop.org \
    --cc=tvrtko.ursulin@linux.intel.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).