From: "Souza, Jose" <jose.souza@intel.com>
To: "intel-gfx@lists.freedesktop.org"
<intel-gfx@lists.freedesktop.org>,
"chris@chris-wilson.co.uk" <chris@chris-wilson.co.uk>
Subject: Re: [Intel-gfx] [PATCH] drm/i915/gt: Make the CTX_TIMESTAMP readable on !rcs
Date: Tue, 2 Jun 2020 19:56:12 +0000 [thread overview]
Message-ID: <78e2807cd9d6d36a8e13ec38f6001ef7413cda1c.camel@intel.com> (raw)
In-Reply-To: <20200602154839.6902-1-chris@chris-wilson.co.uk>
On Tue, 2020-06-02 at 16:48 +0100, Chris Wilson wrote:
> For reasons that be, the HW only allows usersace to read its own
> CTX_TIMESTAMP [context local HW runtime] on rcs. Make it available for
> all by adding it to the whitelists.
>
> v2: The change took effect from Cometlake.
> v3: Ignore timestamps that autoincrement when validating the whitelist
I would have separated add the register to the whitelist from the selftest but anyways looks good.
Reviewed-by: José Roberto de Souza <jose.souza@intel.com>
>
> Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
> ---
> drivers/gpu/drm/i915/gt/intel_workarounds.c | 25 ++++++++++++++++++-
> .../gpu/drm/i915/gt/selftest_workarounds.c | 17 +++++++++++++
> 2 files changed, 41 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/gpu/drm/i915/gt/intel_workarounds.c b/drivers/gpu/drm/i915/gt/intel_workarounds.c
> index 6e1accbcc045..0731bbcef06c 100644
> --- a/drivers/gpu/drm/i915/gt/intel_workarounds.c
> +++ b/drivers/gpu/drm/i915/gt/intel_workarounds.c
> @@ -1206,6 +1206,18 @@ static void cfl_whitelist_build(struct intel_engine_cs *engine)
> RING_FORCE_TO_NONPRIV_RANGE_4);
> }
>
> +static void cml_whitelist_build(struct intel_engine_cs *engine)
> +{
> + struct i915_wa_list *w = &engine->whitelist;
> +
> + if (engine->class != RENDER_CLASS)
> + whitelist_reg_ext(w,
> + RING_CTX_TIMESTAMP(engine->mmio_base),
> + RING_FORCE_TO_NONPRIV_ACCESS_RD);
> +
> + cfl_whitelist_build(engine);
> +}
> +
> static void cnl_whitelist_build(struct intel_engine_cs *engine)
> {
> struct i915_wa_list *w = &engine->whitelist;
> @@ -1256,9 +1268,15 @@ static void icl_whitelist_build(struct intel_engine_cs *engine)
> /* hucStatus2RegOffset */
> whitelist_reg_ext(w, _MMIO(0x23B0 + engine->mmio_base),
> RING_FORCE_TO_NONPRIV_ACCESS_RD);
> + whitelist_reg_ext(w,
> + RING_CTX_TIMESTAMP(engine->mmio_base),
> + RING_FORCE_TO_NONPRIV_ACCESS_RD);
> break;
>
> default:
> + whitelist_reg_ext(w,
> + RING_CTX_TIMESTAMP(engine->mmio_base),
> + RING_FORCE_TO_NONPRIV_ACCESS_RD);
> break;
> }
> }
> @@ -1290,6 +1308,9 @@ static void tgl_whitelist_build(struct intel_engine_cs *engine)
> whitelist_reg(w, HIZ_CHICKEN);
> break;
> default:
> + whitelist_reg_ext(w,
> + RING_CTX_TIMESTAMP(engine->mmio_base),
> + RING_FORCE_TO_NONPRIV_ACCESS_RD);
> break;
> }
> }
> @@ -1307,7 +1328,9 @@ void intel_engine_init_whitelist(struct intel_engine_cs *engine)
> icl_whitelist_build(engine);
> else if (IS_CANNONLAKE(i915))
> cnl_whitelist_build(engine);
> - else if (IS_COFFEELAKE(i915) || IS_COMETLAKE(i915))
> + else if (IS_COMETLAKE(i915))
> + cml_whitelist_build(engine);
> + else if (IS_COFFEELAKE(i915))
> cfl_whitelist_build(engine);
> else if (IS_GEMINILAKE(i915))
> glk_whitelist_build(engine);
> diff --git a/drivers/gpu/drm/i915/gt/selftest_workarounds.c b/drivers/gpu/drm/i915/gt/selftest_workarounds.c
> index 32785463ec9e..febc9e6692ba 100644
> --- a/drivers/gpu/drm/i915/gt/selftest_workarounds.c
> +++ b/drivers/gpu/drm/i915/gt/selftest_workarounds.c
> @@ -417,6 +417,20 @@ static bool wo_register(struct intel_engine_cs *engine, u32 reg)
> return false;
> }
>
> +static bool timestamp(const struct intel_engine_cs *engine, u32 reg)
> +{
> + reg = (reg - engine->mmio_base) & ~RING_FORCE_TO_NONPRIV_ACCESS_MASK;
> + switch (reg) {
> + case 0x358:
> + case 0x35c:
> + case 0x3a8:
> + return true;
> +
> + default:
> + return false;
> + }
> +}
> +
> static bool ro_register(u32 reg)
> {
> if ((reg & RING_FORCE_TO_NONPRIV_ACCESS_MASK) ==
> @@ -497,6 +511,9 @@ static int check_dirty_whitelist(struct intel_context *ce)
> if (wo_register(engine, reg))
> continue;
>
> + if (timestamp(engine, reg))
> + continue; /* timestamps are expected to autoincrement */
> +
> ro_reg = ro_register(reg);
>
> /* Clear non priv flags */
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
next prev parent reply other threads:[~2020-06-02 19:56 UTC|newest]
Thread overview: 13+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-06-02 14:05 [Intel-gfx] [PATCH 1/2] drm/i915: Identify Cometlake platform Chris Wilson
2020-06-02 14:05 ` [Intel-gfx] [PATCH 2/2] drm/i915/gt: Make the CTX_TIMESTAMP readable on !rcs Chris Wilson
2020-06-02 15:48 ` [Intel-gfx] [PATCH] " Chris Wilson
2020-06-02 19:56 ` Souza, Jose [this message]
2020-06-02 15:08 ` [Intel-gfx] ✗ Fi.CI.CHECKPATCH: warning for series starting with [1/2] drm/i915: Identify Cometlake platform Patchwork
2020-06-02 15:09 ` [Intel-gfx] ✗ Fi.CI.SPARSE: " Patchwork
2020-06-02 15:31 ` [Intel-gfx] ✗ Fi.CI.BAT: failure " Patchwork
2020-06-02 16:27 ` [Intel-gfx] ✗ Fi.CI.CHECKPATCH: warning for series starting with [1/2] drm/i915: Identify Cometlake platform (rev2) Patchwork
2020-06-02 16:28 ` [Intel-gfx] ✗ Fi.CI.SPARSE: " Patchwork
2020-06-02 16:48 ` [Intel-gfx] ✓ Fi.CI.BAT: success " Patchwork
2020-06-02 19:52 ` [Intel-gfx] [PATCH 1/2] drm/i915: Identify Cometlake platform Souza, Jose
2020-06-03 0:35 ` [Intel-gfx] ✓ Fi.CI.IGT: success for series starting with [1/2] drm/i915: Identify Cometlake platform (rev2) Patchwork
-- strict thread matches above, loose matches on Subject: below --
2020-06-02 9:59 [Intel-gfx] [PATCH] drm/i915/gt: Make the CTX_TIMESTAMP readable on !rcs Chris Wilson
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=78e2807cd9d6d36a8e13ec38f6001ef7413cda1c.camel@intel.com \
--to=jose.souza@intel.com \
--cc=chris@chris-wilson.co.uk \
--cc=intel-gfx@lists.freedesktop.org \
/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).