All of lore.kernel.org
 help / color / mirror / Atom feed
From: Jan Beulich <jbeulich@suse.com>
To: "xen-devel@lists.xenproject.org" <xen-devel@lists.xenproject.org>
Cc: "Andrew Cooper" <andrew.cooper3@citrix.com>,
	"Roger Pau Monné" <roger.pau@citrix.com>
Subject: [PATCH v9 01/10] x86emul: support LKGS
Date: Mon, 24 Nov 2025 15:57:52 +0100	[thread overview]
Message-ID: <451ca993-86f2-40dc-b68b-a23c055687c0@suse.com> (raw)
In-Reply-To: <926a2315-a2b7-4aad-87e6-d686c9da9e3a@suse.com>

Provide support for this insn, which is a prereq to FRED. CPUID-wise,
while its and FRED's enumerators were already introduced, their dependency
still needs adding.

While adding a testcase, also add a SWAPGS one. In order to not affect
the behavior of pre-existing tests, install write_{segment,msr} hooks
only transiently.

Signed-off-by: Jan Beulich <jbeulich@suse.com>
---
Instead of ->read_segment() we could of course also use ->read_msr() to
fetch the original GS base. I don't think I can see a clear advantage of
either approach; the way it's done it matches how we handle SWAPGS.

For PV save_segments() would need adjustment, but the insn being
restricted to ring 0 means PV guests can't use it anyway (unless we
wanted to emulate it as another privileged insn).
---
v9: Re-base.
v8: Re-base.
v6: Use MSR constants in test harness. S->s in cpufeatureset.h. Add
    NMI_SRC feature bits. Re-base.
v5: Re-base.
v3: Add dependency on LM. Re-base.
v2: Use X86_EXC_*. Add comments.

--- a/tools/tests/x86_emulator/predicates.c
+++ b/tools/tests/x86_emulator/predicates.c
@@ -326,6 +326,7 @@ static const struct {
     { { 0x00, 0x18 }, { 2, 2 }, T, R }, /* ltr */
     { { 0x00, 0x20 }, { 2, 2 }, T, R }, /* verr */
     { { 0x00, 0x28 }, { 2, 2 }, T, R }, /* verw */
+    { { 0x00, 0x30 }, { 0, 2 }, T, R, pfx_f2 }, /* lkgs */
     { { 0x01, 0x00 }, { 2, 2 }, F, W }, /* sgdt */
     { { 0x01, 0x08 }, { 2, 2 }, F, W }, /* sidt */
     { { 0x01, 0x10 }, { 2, 2 }, F, R }, /* lgdt */
--- a/tools/tests/x86_emulator/test_x86_emulator.c
+++ b/tools/tests/x86_emulator/test_x86_emulator.c
@@ -672,6 +672,10 @@ static int blk(
     return x86_emul_blk((void *)offset, p_data, bytes, eflags, state, ctxt);
 }
 
+#ifdef __x86_64__
+static unsigned long gs_base, gs_base_shadow;
+#endif
+
 static int read_segment(
     enum x86_segment seg,
     struct segment_register *reg,
@@ -681,8 +685,30 @@ static int read_segment(
         return X86EMUL_UNHANDLEABLE;
     memset(reg, 0, sizeof(*reg));
     reg->p = 1;
+
+#ifdef __x86_64__
+    if ( seg == x86_seg_gs )
+        reg->base = gs_base;
+#endif
+
+    return X86EMUL_OKAY;
+}
+
+#ifdef __x86_64__
+static int write_segment(
+    enum x86_segment seg,
+    const struct segment_register *reg,
+    struct x86_emulate_ctxt *ctxt)
+{
+    if ( !is_x86_user_segment(seg) )
+        return X86EMUL_UNHANDLEABLE;
+
+    if ( seg == x86_seg_gs )
+        gs_base = reg->base;
+
     return X86EMUL_OKAY;
 }
+#endif
 
 static int read_msr(
     unsigned int reg,
@@ -695,6 +721,20 @@ static int read_msr(
         *val = ctxt->addr_size > 32 ? EFER_LME | EFER_LMA : 0;
         return X86EMUL_OKAY;
 
+#ifdef __x86_64__
+    case MSR_GS_BASE:
+        if ( ctxt->addr_size < 64 )
+            break;
+        *val = gs_base;
+        return X86EMUL_OKAY;
+
+    case MSR_SHADOW_GS_BASE:
+        if ( ctxt->addr_size < 64 )
+            break;
+        *val = gs_base_shadow;
+        return X86EMUL_OKAY;
+#endif
+
     case MSR_TSC_AUX:
 #define TSC_AUX_VALUE 0xCACACACA
         *val = TSC_AUX_VALUE;
@@ -704,6 +744,31 @@ static int read_msr(
     return X86EMUL_UNHANDLEABLE;
 }
 
+#ifdef __x86_64__
+static int write_msr(
+    unsigned int reg,
+    uint64_t val,
+    struct x86_emulate_ctxt *ctxt)
+{
+    switch ( reg )
+    {
+    case MSR_GS_BASE:
+        if ( ctxt->addr_size < 64 || !is_canonical_address(val) )
+            break;
+        gs_base = val;
+        return X86EMUL_OKAY;
+
+    case MSR_SHADOW_GS_BASE:
+        if ( ctxt->addr_size < 64 || !is_canonical_address(val) )
+            break;
+        gs_base_shadow = val;
+        return X86EMUL_OKAY;
+    }
+
+    return X86EMUL_UNHANDLEABLE;
+}
+#endif
+
 #define INVPCID_ADDR 0x12345678
 #define INVPCID_PCID 0x123
 
@@ -1338,6 +1403,41 @@ int main(int argc, char **argv)
         printf("%u bytes read - ", bytes_read);
         goto fail;
     }
+    printf("okay\n");
+
+    emulops.write_segment = write_segment;
+    emulops.write_msr     = write_msr;
+
+    printf("%-40s", "Testing swapgs...");
+    instr[0] = 0x0f; instr[1] = 0x01; instr[2] = 0xf8;
+    regs.eip = (unsigned long)&instr[0];
+    gs_base = 0xffffeeeecccc8888UL;
+    gs_base_shadow = 0x0000111122224444UL;
+    rc = x86_emulate(&ctxt, &emulops);
+    if ( (rc != X86EMUL_OKAY) ||
+         (regs.eip != (unsigned long)&instr[3]) ||
+         (gs_base != 0x0000111122224444UL) ||
+         (gs_base_shadow != 0xffffeeeecccc8888UL) )
+        goto fail;
+    printf("okay\n");
+
+    printf("%-40s", "Testing lkgs 2(%rdx)...");
+    instr[0] = 0xf2; instr[1] = 0x0f; instr[2] = 0x00; instr[3] = 0x72; instr[4] = 0x02;
+    regs.eip = (unsigned long)&instr[0];
+    regs.edx = (unsigned long)res;
+    res[0]   = 0x00004444;
+    res[1]   = 0x8888cccc;
+    i = cpu_policy.extd.nscb; cpu_policy.extd.nscb = true; /* for AMD */
+    rc = x86_emulate(&ctxt, &emulops);
+    if ( (rc != X86EMUL_OKAY) ||
+         (regs.eip != (unsigned long)&instr[5]) ||
+         (gs_base != 0x0000111122224444UL) ||
+         gs_base_shadow )
+        goto fail;
+
+    cpu_policy.extd.nscb = i;
+    emulops.write_segment = NULL;
+    emulops.write_msr     = NULL;
 #endif
     printf("okay\n");
 
--- a/tools/tests/x86_emulator/x86-emulate.c
+++ b/tools/tests/x86_emulator/x86-emulate.c
@@ -85,6 +85,7 @@ bool emul_test_init(void)
     cpu_policy.feat.invpcid = true;
     cpu_policy.feat.adx = true;
     cpu_policy.feat.rdpid = true;
+    cpu_policy.feat.lkgs = true;
     cpu_policy.feat.wrmsrns = true;
     cpu_policy.extd.clzero = true;
 
--- a/xen/arch/x86/x86_emulate/decode.c
+++ b/xen/arch/x86/x86_emulate/decode.c
@@ -744,8 +744,12 @@ decode_twobyte(struct x86_emulate_state
         case 0:
             s->desc |= DstMem | SrcImplicit | Mov;
             break;
+        case 6:
+            if ( !(s->modrm_reg & 1) && mode_64bit() )
+            {
         case 2: case 4:
-            s->desc |= SrcMem16;
+                s->desc |= SrcMem16;
+            }
             break;
         }
         break;
--- a/xen/arch/x86/x86_emulate/private.h
+++ b/xen/arch/x86/x86_emulate/private.h
@@ -608,6 +608,7 @@ amd_like(const struct x86_emulate_ctxt *
 #define vcpu_has_avx_vnni()    (ctxt->cpuid->feat.avx_vnni)
 #define vcpu_has_avx512_bf16() (ctxt->cpuid->feat.avx512_bf16)
 #define vcpu_has_cmpccxadd()   (ctxt->cpuid->feat.cmpccxadd)
+#define vcpu_has_lkgs()        (ctxt->cpuid->feat.lkgs)
 #define vcpu_has_wrmsrns()     (ctxt->cpuid->feat.wrmsrns)
 #define vcpu_has_avx_ifma()    (ctxt->cpuid->feat.avx_ifma)
 #define vcpu_has_avx_vnni_int8() (ctxt->cpuid->feat.avx_vnni_int8)
--- a/xen/arch/x86/x86_emulate/x86_emulate.c
+++ b/xen/arch/x86/x86_emulate/x86_emulate.c
@@ -2899,8 +2899,35 @@ x86_emulate(
                 break;
             }
             break;
-        default:
-            generate_exception_if(true, X86_EXC_UD);
+        case 6: /* lkgs */
+            generate_exception_if((modrm_reg & 1) || vex.pfx != vex_f2,
+                                  X86_EXC_UD);
+            generate_exception_if(!mode_64bit() || !mode_ring0(), X86_EXC_UD);
+            vcpu_must_have(lkgs);
+            fail_if(!ops->read_segment || !ops->read_msr ||
+                    !ops->write_segment || !ops->write_msr);
+            if ( (rc = ops->read_msr(MSR_SHADOW_GS_BASE, &msr_val,
+                                     ctxt)) != X86EMUL_OKAY ||
+                 (rc = ops->read_segment(x86_seg_gs, &sreg,
+                                         ctxt)) != X86EMUL_OKAY )
+                goto done;
+            dst.orig_val = sreg.base; /* Preserve full GS Base. */
+            if ( (rc = protmode_load_seg(x86_seg_gs, src.val, false, &sreg,
+                                         ctxt, ops)) != X86EMUL_OKAY ||
+                 /* Write (32-bit) base into SHADOW_GS. */
+                 (rc = ops->write_msr(MSR_SHADOW_GS_BASE, sreg.base,
+                                      ctxt)) != X86EMUL_OKAY )
+                goto done;
+            sreg.base = dst.orig_val; /* Reinstate full GS Base. */
+            if ( (rc = ops->write_segment(x86_seg_gs, &sreg,
+                                          ctxt)) != X86EMUL_OKAY )
+            {
+                /* Best effort unwind (i.e. no real error checking). */
+                if ( ops->write_msr(MSR_SHADOW_GS_BASE, msr_val,
+                                    ctxt) == X86EMUL_EXCEPTION )
+                    x86_emul_reset_event(ctxt);
+                goto done;
+            }
             break;
         }
         break;
--- a/xen/tools/gen-cpuid.py
+++ b/xen/tools/gen-cpuid.py
@@ -282,7 +282,8 @@ def crunch_numbers(state):
         # superpages, PCID and PKU are only available in 4 level paging.
         # NO_LMSL indicates the absense of Long Mode Segment Limits, which
         # have been dropped in hardware.
-        LM: [CX16, PCID, LAHF_LM, PAGE1GB, PKU, NO_LMSL, AMX_TILE, CMPCCXADD],
+        LM: [CX16, PCID, LAHF_LM, PAGE1GB, PKU, NO_LMSL, AMX_TILE, CMPCCXADD,
+             LKGS],
 
         # AMD K6-2+ and K6-III processors shipped with 3DNow+, beyond the
         # standard 3DNow in the earlier K6 processors.
@@ -351,6 +352,9 @@ def crunch_numbers(state):
         # computational instructions.  All further AMX features are built on top
         # of AMX-TILE.
         AMX_TILE: [AMX_BF16, AMX_INT8, AMX_FP16, AMX_COMPLEX],
+
+        # FRED builds on the LKGS instruction.
+        LKGS: [FRED],
     }
 
     deep_features = tuple(sorted(deps.keys()))



  reply	other threads:[~2025-11-24 14:58 UTC|newest]

Thread overview: 26+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-11-24 14:56 [PATCH v9 10/10] x86emul: misc additions Jan Beulich
2025-11-24 14:57 ` Jan Beulich [this message]
2025-11-24 14:58 ` [PATCH v9 02/10] x86emul+VMX: support {RD,WR}MSRLIST Jan Beulich
2025-11-24 14:58 ` [PATCH v9 03/10] x86emul: support USER_MSR instructions Jan Beulich
2025-11-24 14:59 ` [PATCH v9 04/10] x86/cpu-policy: re-arrange no-VMX logic Jan Beulich
2026-04-07 21:58   ` Andrew Cooper
2026-04-08  6:09     ` Jan Beulich
2025-11-24 15:00 ` [PATCH v9 05/10] VMX: support USER-MSR Jan Beulich
2025-11-24 15:00 ` [PATCH v9 06/10] x86emul: support MSR_IMM instructions Jan Beulich
2025-11-24 15:00 ` [PATCH v9 07/10] VMX: support MSR-IMM Jan Beulich
2025-11-26 18:50   ` Andrew Cooper
2025-11-27  8:18     ` Jan Beulich
2025-11-24 15:01 ` [PATCH v9 08/10] x86emul: support non-SIMD MOVRS Jan Beulich
2025-11-24 15:01 ` [PATCH v9 09/10] x86: use / "support" UDB Jan Beulich
2025-12-05 12:01   ` Andrew Cooper
2025-12-05 12:40     ` Andrew Cooper
2025-12-05 13:13       ` Jan Beulich
2025-12-05 13:15         ` Andrew Cooper
2025-12-05 13:15       ` Jan Beulich
2025-12-05 13:35         ` Andrew Cooper
2025-12-05 13:09     ` Jan Beulich
2025-11-24 15:02 ` [PATCH v9 10/10] x86emul: support AVX512-BMM Jan Beulich
2025-12-05 12:33   ` Andrew Cooper
2025-12-05 12:47     ` Jan Beulich
2026-04-07 15:11       ` Andrew Cooper
2025-11-24 15:03 ` [PATCH v9 00/10] x86emul: misc additions 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=451ca993-86f2-40dc-b68b-a23c055687c0@suse.com \
    --to=jbeulich@suse.com \
    --cc=andrew.cooper3@citrix.com \
    --cc=roger.pau@citrix.com \
    --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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.