xen-devel.lists.xenproject.org archive mirror
 help / color / mirror / Atom feed
From: Andrew Cooper <andrew.cooper3@citrix.com>
To: Xen-devel <xen-devel@lists.xenproject.org>
Cc: "Andrew Cooper" <andrew.cooper3@citrix.com>,
	"Jan Beulich" <JBeulich@suse.com>,
	"Roger Pau Monné" <roger.pau@citrix.com>, "Wei Liu" <wl@xen.org>
Subject: [PATCH 4/5] x86/cpuid: Simplify recalculate_xstate()
Date: Mon, 3 May 2021 16:39:37 +0100	[thread overview]
Message-ID: <20210503153938.14109-5-andrew.cooper3@citrix.com> (raw)
In-Reply-To: <20210503153938.14109-1-andrew.cooper3@citrix.com>

Make use of the new xstate_uncompressed_size() helper rather than maintaining
the running calculation while accumulating feature components.

The rest of the CPUID data can come direct from the raw cpuid policy.  All
per-component data forms an ABI through the behaviour of the X{SAVE,RSTOR}*
instructions, and are constant.

Use for_each_set_bit() rather than opencoding a slightly awkward version of
it.  Mask the attributes in ecx down based on the visible features.  This
isn't actually necessary for any components or attributes defined at the time
of writing (up to AMX), but is added out of an abundance of caution.

Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
---
CC: Jan Beulich <JBeulich@suse.com>
CC: Roger Pau Monné <roger.pau@citrix.com>
CC: Wei Liu <wl@xen.org>

Using min() in for_each_set_bit() leads to awful code generation, as it
prohibits the optimiations for spotting that the bitmap is <= BITS_PER_LONG.
As p->xstate is long enough already, use a BUILD_BUG_ON() instead.
---
 xen/arch/x86/cpuid.c | 52 +++++++++++++++++-----------------------------------
 1 file changed, 17 insertions(+), 35 deletions(-)

diff --git a/xen/arch/x86/cpuid.c b/xen/arch/x86/cpuid.c
index 752bf244ea..c7f8388e5d 100644
--- a/xen/arch/x86/cpuid.c
+++ b/xen/arch/x86/cpuid.c
@@ -154,8 +154,7 @@ static void sanitise_featureset(uint32_t *fs)
 static void recalculate_xstate(struct cpuid_policy *p)
 {
     uint64_t xstates = XSTATE_FP_SSE;
-    uint32_t xstate_size = XSTATE_AREA_MIN_SIZE;
-    unsigned int i, Da1 = p->xstate.Da1;
+    unsigned int i, ecx_bits = 0, Da1 = p->xstate.Da1;
 
     /*
      * The Da1 leaf is the only piece of information preserved in the common
@@ -167,61 +166,44 @@ static void recalculate_xstate(struct cpuid_policy *p)
         return;
 
     if ( p->basic.avx )
-    {
         xstates |= X86_XCR0_YMM;
-        xstate_size = max(xstate_size,
-                          xstate_offsets[X86_XCR0_YMM_POS] +
-                          xstate_sizes[X86_XCR0_YMM_POS]);
-    }
 
     if ( p->feat.mpx )
-    {
         xstates |= X86_XCR0_BNDREGS | X86_XCR0_BNDCSR;
-        xstate_size = max(xstate_size,
-                          xstate_offsets[X86_XCR0_BNDCSR_POS] +
-                          xstate_sizes[X86_XCR0_BNDCSR_POS]);
-    }
 
     if ( p->feat.avx512f )
-    {
         xstates |= X86_XCR0_OPMASK | X86_XCR0_ZMM | X86_XCR0_HI_ZMM;
-        xstate_size = max(xstate_size,
-                          xstate_offsets[X86_XCR0_HI_ZMM_POS] +
-                          xstate_sizes[X86_XCR0_HI_ZMM_POS]);
-    }
 
     if ( p->feat.pku )
-    {
         xstates |= X86_XCR0_PKRU;
-        xstate_size = max(xstate_size,
-                          xstate_offsets[X86_XCR0_PKRU_POS] +
-                          xstate_sizes[X86_XCR0_PKRU_POS]);
-    }
 
-    p->xstate.max_size  =  xstate_size;
+    /* Subleaf 0 */
+    p->xstate.max_size =
+        xstate_uncompressed_size(xstates & ~XSTATE_XSAVES_ONLY);
     p->xstate.xcr0_low  =  xstates & ~XSTATE_XSAVES_ONLY;
     p->xstate.xcr0_high = (xstates & ~XSTATE_XSAVES_ONLY) >> 32;
 
+    /* Subleaf 1 */
     p->xstate.Da1 = Da1;
     if ( p->xstate.xsaves )
     {
+        ecx_bits |= 3; /* Align64, XSS */
         p->xstate.xss_low   =  xstates & XSTATE_XSAVES_ONLY;
         p->xstate.xss_high  = (xstates & XSTATE_XSAVES_ONLY) >> 32;
     }
-    else
-        xstates &= ~XSTATE_XSAVES_ONLY;
 
-    for ( i = 2; i < min(63ul, ARRAY_SIZE(p->xstate.comp)); ++i )
+    /* Subleafs 2+ */
+    xstates &= ~XSTATE_FP_SSE;
+    BUILD_BUG_ON(ARRAY_SIZE(p->xstate.comp) < 63);
+    for_each_set_bit ( i, &xstates, 63 )
     {
-        uint64_t curr_xstate = 1ul << i;
-
-        if ( !(xstates & curr_xstate) )
-            continue;
-
-        p->xstate.comp[i].size   = xstate_sizes[i];
-        p->xstate.comp[i].offset = xstate_offsets[i];
-        p->xstate.comp[i].xss    = curr_xstate & XSTATE_XSAVES_ONLY;
-        p->xstate.comp[i].align  = curr_xstate & xstate_align;
+        /*
+         * Pass through size (eax) and offset (ebx) directly.  Visbility of
+         * attributes in ecx limited by visible features in Da1.
+         */
+        p->xstate.raw[i].a = raw_cpuid_policy.xstate.raw[i].a;
+        p->xstate.raw[i].b = raw_cpuid_policy.xstate.raw[i].b;
+        p->xstate.raw[i].c = raw_cpuid_policy.xstate.raw[i].c & ecx_bits;
     }
 }
 
-- 
2.11.0



  parent reply	other threads:[~2021-05-03 15:40 UTC|newest]

Thread overview: 24+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-05-03 15:39 [PATCH 0/5] x86/xstate: Fixes to size calculations Andrew Cooper
2021-05-03 15:39 ` [PATCH 1/5] x86/xstate: Elide redundant writes in set_xcr0() Andrew Cooper
2021-05-04 11:51   ` Jan Beulich
2021-05-03 15:39 ` [PATCH 2/5] x86/xstate: Rename _xstate_ctxt_size() to hw_uncompressed_size() Andrew Cooper
2021-05-04 11:53   ` Jan Beulich
2021-05-03 15:39 ` [PATCH 3/5] x86/xstate: Rework xstate_ctxt_size() as xstate_uncompressed_size() Andrew Cooper
2021-05-03 18:17   ` Andrew Cooper
2021-05-04 12:08     ` Jan Beulich
2021-05-04 12:15       ` Andrew Cooper
2021-05-04 12:20   ` Jan Beulich
2021-05-04 12:22     ` Andrew Cooper
2021-05-04 12:45       ` Jan Beulich
2021-05-03 15:39 ` Andrew Cooper [this message]
2021-05-04 12:43   ` [PATCH 4/5] x86/cpuid: Simplify recalculate_xstate() Jan Beulich
2021-05-04 13:58     ` Andrew Cooper
2021-05-05  8:19       ` Jan Beulich
2021-05-05 14:53         ` Andrew Cooper
2021-05-05 15:14           ` Jan Beulich
2021-05-03 15:39 ` [PATCH 5/5] x86/cpuid: Fix handling of xsave dynamic leaves Andrew Cooper
2021-05-04 12:56   ` Jan Beulich
2021-05-04 14:17     ` Andrew Cooper
2021-05-05  8:33       ` Jan Beulich
2021-05-05 16:59         ` Andrew Cooper
2021-05-06  6:17           ` Jan Beulich

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=20210503153938.14109-5-andrew.cooper3@citrix.com \
    --to=andrew.cooper3@citrix.com \
    --cc=JBeulich@suse.com \
    --cc=roger.pau@citrix.com \
    --cc=wl@xen.org \
    --cc=xen-devel@lists.xenproject.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).