All of lore.kernel.org
 help / color / mirror / Atom feed
From: Igor Mammedov <imammedo@redhat.com>
To: Eduardo Habkost <ehabkost@redhat.com>
Cc: qemu-devel@nongnu.org
Subject: Re: [Qemu-devel] [PATCH 1/2] target-i386: cpu: convert plus/minus properties to global properties
Date: Mon, 21 Aug 2017 10:32:41 +0200	[thread overview]
Message-ID: <20170821103241.2651a5bc@nial.brq.redhat.com> (raw)
In-Reply-To: <20170818174029.GA27715@localhost.localdomain>

On Fri, 18 Aug 2017 14:40:29 -0300
Eduardo Habkost <ehabkost@redhat.com> wrote:

> On Thu, Aug 17, 2017 at 04:07:56PM +0200, Igor Mammedov wrote:
> > Since
> >  (commit d4a606b3 i386: Don't override -cpu options on -cpu host/max)
> > it became possible to delete hack where it was necessary to
> > postpone applying plus/minus features to realize time
> > after max_features were applied to keep legacy +-feat
> > override semantics.
> > 
> > With above commit it's possible to convert +-feat to a set
> > of GlobalProperty items and keep +-feat override semantics,
> > these properties should be added to global list at the end
> > to override properties that were set with feat=on|off syntax.
> > 
> > Signed-off-by: Igor Mammedov <imammedo@redhat.com>  
> [...]
> > +/* Parse "+feature,-feature,feature=foo" CPU feature string */
> >  static void x86_cpu_parse_featurestr(const char *typename, char *features,
> >                                       Error **errp)
> >  {
> > +    /* Compatibily hack to maintain legacy +-feat semantic,
> > +     * where +-feat overwrites any feature set by
> > +     * feat=on|feat even if the later is parsed after +-feat
> > +     * (i.e. "-x2apic,x2apic=on" will result in x2apic disabled)
> > +     */
> > +    GList *l, *plus_features = NULL, *minus_features = NULL;  
> 
> The warning about ambiguous CPU options exists since 2.8, I think
> this is a good opportunity to get rid of the "[+-]feat overrides
> feat=on|off" rule and simplify the parsing code.  Do you want to
> do this in the same patch?
I'd prefer not to do it in this patch/series, as it's not related.

WE can do cleanups later on top.

> 
> 
> >      char *featurestr; /* Single 'key=value" string being parsed */
> >      static bool cpu_globals_initialized;
> >      bool ambiguous = false;
> > @@ -2095,7 +2104,6 @@ static void x86_cpu_parse_featurestr(const char *typename, char *features,
> >          const char *val = NULL;
> >          char *eq = NULL;
> >          char num[32];
> > -        GlobalProperty *prop;
> >  
> >          /* Compatibility syntax: */
> >          if (featurestr[0] == '+') {
> > @@ -2147,21 +2155,32 @@ static void x86_cpu_parse_featurestr(const char *typename, char *features,
> >              name = "tsc-frequency";
> >          }
> >  
> > -        prop = g_new0(typeof(*prop), 1);
> > -        prop->driver = typename;
> > -        prop->property = g_strdup(name);
> > -        prop->value = g_strdup(val);
> > -        prop->errp = &error_fatal;
> > -        qdev_prop_register_global(prop);
> > +        cpu_add_feat_as_prop(typename, name, val);
> >      }
> >  
> >      if (ambiguous) {
> >          warn_report("Compatibility of ambiguous CPU model "
> >                      "strings won't be kept on future QEMU versions");
> >      }
> > +
> > +    for (l = plus_features; l; l = l->next) {
> > +        const char *name = l->data;
> > +        cpu_add_feat_as_prop(typename, name, "on");
> > +    }
> > +    if (plus_features) {
> > +        g_list_free_full(plus_features, g_free);
> > +    }
> > +
> > +    for (l = minus_features; l; l = l->next) {
> > +        const char *name = l->data;
> > +        cpu_add_feat_as_prop(typename, name, "off");
> > +    }
> > +    if (minus_features) {
> > +        g_list_free_full(minus_features, g_free);
> > +    }
> >  }
> >  
> > -static void x86_cpu_expand_features(X86CPU *cpu, Error **errp);
> > +static void x86_cpu_expand_features(X86CPU *cpu);
> >  static int x86_cpu_filter_features(X86CPU *cpu);
> >  
> >  /* Check for missing features that may prevent the CPU class from
> > @@ -2172,7 +2191,6 @@ static void x86_cpu_class_check_missing_features(X86CPUClass *xcc,
> >  {
> >      X86CPU *xc;
> >      FeatureWord w;
> > -    Error *err = NULL;
> >      strList **next = missing_feats;
> >  
> >      if (xcc->kvm_required && !kvm_enabled()) {
> > @@ -2184,18 +2202,7 @@ static void x86_cpu_class_check_missing_features(X86CPUClass *xcc,
> >  
> >      xc = X86_CPU(object_new(object_class_get_name(OBJECT_CLASS(xcc))));
> >  
> > -    x86_cpu_expand_features(xc, &err);
> > -    if (err) {
> > -        /* Errors at x86_cpu_expand_features should never happen,
> > -         * but in case it does, just report the model as not
> > -         * runnable at all using the "type" property.
> > -         */
> > -        strList *new = g_new0(strList, 1);
> > -        new->value = g_strdup("type");
> > -        *next = new;
> > -        next = &new->next;
> > -    }
> > -
> > +    x86_cpu_expand_features(xc);
> >      x86_cpu_filter_features(xc);
> >  
> >      for (w = 0; w < FEATURE_WORDS; w++) {
> > @@ -2559,11 +2566,7 @@ static X86CPU *x86_cpu_from_model(const char *model, QDict *props, Error **errp)
> >          }
> >      }
> >  
> > -    x86_cpu_expand_features(xc, &err);
> > -    if (err) {
> > -        goto out;
> > -    }
> > -
> > +    x86_cpu_expand_features(xc);
> >  out:
> >      if (err) {
> >          error_propagate(errp, err);
> > @@ -3453,18 +3456,11 @@ static void x86_cpu_enable_xsave_components(X86CPU *cpu)
> >  /* Expand CPU configuration data, based on configured features
> >   * and host/accelerator capabilities when appropriate.
> >   */
> > -static void x86_cpu_expand_features(X86CPU *cpu, Error **errp)
> > +static void x86_cpu_expand_features(X86CPU *cpu)
> >  {
> >      CPUX86State *env = &cpu->env;
> >      FeatureWord w;
> > -    GList *l;
> > -    Error *local_err = NULL;
> >  
> > -    /*TODO: Now cpu->max_features doesn't overwrite features
> > -     * set using QOM properties, and we can convert
> > -     * plus_features & minus_features to global properties
> > -     * inside x86_cpu_parse_featurestr() too.
> > -     */
> >      if (cpu->max_features) {
> >          for (w = 0; w < FEATURE_WORDS; w++) {
> >              /* Override only features that weren't set explicitly
> > @@ -3476,22 +3472,6 @@ static void x86_cpu_expand_features(X86CPU *cpu, Error **errp)
> >          }
> >      }
> >  
> > -    for (l = plus_features; l; l = l->next) {
> > -        const char *prop = l->data;
> > -        object_property_set_bool(OBJECT(cpu), true, prop, &local_err);
> > -        if (local_err) {
> > -            goto out;
> > -        }
> > -    }
> > -
> > -    for (l = minus_features; l; l = l->next) {
> > -        const char *prop = l->data;
> > -        object_property_set_bool(OBJECT(cpu), false, prop, &local_err);
> > -        if (local_err) {
> > -            goto out;
> > -        }
> > -    }
> > -
> >      if (!kvm_enabled() || !cpu->expose_kvm) {
> >          env->features[FEAT_KVM] = 0;
> >      }
> > @@ -3527,11 +3507,6 @@ static void x86_cpu_expand_features(X86CPU *cpu, Error **errp)
> >      if (env->cpuid_xlevel2 == UINT32_MAX) {
> >          env->cpuid_xlevel2 = env->cpuid_min_xlevel2;
> >      }
> > -
> > -out:
> > -    if (local_err != NULL) {
> > -        error_propagate(errp, local_err);
> > -    }
> >  }
> >  
> >  /*
> > @@ -3587,10 +3562,7 @@ static void x86_cpu_realizefn(DeviceState *dev, Error **errp)
> >          return;
> >      }
> >  
> > -    x86_cpu_expand_features(cpu, &local_err);
> > -    if (local_err) {
> > -        goto out;
> > -    }
> > +    x86_cpu_expand_features(cpu);
> >  
> >      if (x86_cpu_filter_features(cpu) &&
> >          (cpu->check_cpuid || cpu->enforce_cpuid)) {
> > -- 
> > 2.7.4
> > 
> >   
> 

  reply	other threads:[~2017-08-21  8:42 UTC|newest]

Thread overview: 86+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2017-07-14 13:51 [Qemu-devel] [PATCH 00/28] complete cpu QOMification and remove cpu_FOO_init() helpers Igor Mammedov
2017-07-14 13:51 ` [Qemu-devel] [PATCH 01/28] mips: cpu: move mmu/fpu/mvp_init to realize time Igor Mammedov
2017-07-15 21:48   ` Philippe Mathieu-Daudé
2017-07-14 13:51 ` [Qemu-devel] [PATCH 02/28] mips: MIPSCPU model subclasses Igor Mammedov
2017-07-15 21:48   ` Philippe Mathieu-Daudé
2017-08-17  3:38     ` Philippe Mathieu-Daudé
2017-08-17 10:53       ` Igor Mammedov
2017-08-17 11:15         ` Philippe Mathieu-Daudé
2017-07-14 13:51 ` [Qemu-devel] [PATCH 03/28] mips: replace cpu_mips_init() with cpu_generic_init() Igor Mammedov
2017-07-15  6:09   ` Hervé Poussineau
2017-07-15 21:48   ` Philippe Mathieu-Daudé
2017-07-14 13:51 ` [Qemu-devel] [PATCH 04/28] sparc: convert cpu models to SPARC cpu subclasses Igor Mammedov
2017-08-14  7:56   ` Igor Mammedov
2017-08-14 16:24     ` Artyom Tarasenko
2017-08-15  7:38       ` Igor Mammedov
2017-08-15 11:27     ` Mark Cave-Ayland
2017-08-17  3:50   ` Philippe Mathieu-Daudé
2017-08-17 14:11     ` Igor Mammedov
2017-07-14 13:51 ` [Qemu-devel] [PATCH 05/28] sparc: embed sparc_def_t into CPUSPARCState Igor Mammedov
2017-07-14 13:51 ` [Qemu-devel] [PATCH 06/28] sparc: convert cpu features to qdev properties Igor Mammedov
2017-07-14 13:51 ` [Qemu-devel] [PATCH 07/28] sparc: move adhoc CPUSPARCState initialization to realize time Igor Mammedov
2017-07-14 13:51 ` [Qemu-devel] [PATCH 08/28] x86: extract legacy cpu features format parser Igor Mammedov
2017-08-16 19:32   ` Eduardo Habkost
2017-08-17 14:07     ` [Qemu-devel] [PATCH 1/2] target-i386: cpu: convert plus/minus properties to global properties Igor Mammedov
2017-08-17 14:07       ` [Qemu-devel] [PATCH 2/2] x86: extract legacy cpu features format parser Igor Mammedov
2017-08-18 17:40       ` [Qemu-devel] [PATCH 1/2] target-i386: cpu: convert plus/minus properties to global properties Eduardo Habkost
2017-08-21  8:32         ` Igor Mammedov [this message]
2017-08-23 14:24           ` Eduardo Habkost
2017-08-23 15:54             ` Igor Mammedov
2017-08-23 16:52               ` Eduardo Habkost
2017-07-14 13:52 ` [Qemu-devel] [PATCH 09/28] sparc: replace custom cpu feature parsing with cpu_legacy_parse_featurestr() Igor Mammedov
2017-07-14 13:52 ` [Qemu-devel] [PATCH 10/28] sparc: replace cpu_sparc_init() with cpu_generic_init() Igor Mammedov
2017-07-14 13:52 ` [Qemu-devel] [PATCH 11/28] s390x: replace cpu_s390x_init() " Igor Mammedov
2017-07-18 12:30   ` Cornelia Huck
2017-07-18 13:17     ` Igor Mammedov
2017-08-14  8:03       ` Igor Mammedov
2017-08-14  8:53         ` Cornelia Huck
2017-08-14  9:24           ` Igor Mammedov
2017-08-14  9:27             ` Cornelia Huck
2017-07-14 13:52 ` [Qemu-devel] [PATCH 12/28] alpha: replace cpu_alpha_init() " Igor Mammedov
2017-07-15 18:05   ` Richard Henderson
2017-07-14 13:52 ` [Qemu-devel] [PATCH 13/28] hppa: replace cpu_hppa_init() " Igor Mammedov
2017-07-15 18:06   ` Richard Henderson
2017-07-14 13:52 ` [Qemu-devel] [PATCH 14/28] m68k: replace cpu_m68k_init() " Igor Mammedov
2017-07-15  8:05   ` Thomas Huth
2017-07-15 18:08   ` Richard Henderson
2017-07-15 20:57     ` Laurent Vivier
2017-07-17 10:41     ` Igor Mammedov
2017-07-17 15:05       ` Andreas Färber
2017-07-17 15:23         ` Igor Mammedov
2017-08-14  8:00           ` Igor Mammedov
2017-08-14 18:23             ` Laurent Vivier
2017-07-14 13:52 ` [Qemu-devel] [PATCH 15/28] microblaze: replace cpu_mb_init() " Igor Mammedov
2017-07-15 21:51   ` Philippe Mathieu-Daudé
2017-07-14 13:52 ` [Qemu-devel] [PATCH 16/28] nios2: replace cpu_nios2_init() " Igor Mammedov
2017-07-15 21:53   ` Philippe Mathieu-Daudé
2017-07-14 13:52 ` [Qemu-devel] [PATCH 17/28] tilegx: replace cpu_tilegx_init() " Igor Mammedov
2017-08-16 19:53   ` Eduardo Habkost
2017-07-14 13:52 ` [Qemu-devel] [PATCH 18/28] xtensa: replace cpu_xtensa_init() " Igor Mammedov
2017-08-16 19:56   ` Eduardo Habkost
2017-08-17 14:32     ` Igor Mammedov
2017-08-18 16:50       ` Eduardo Habkost
2017-07-14 13:52 ` [Qemu-devel] [PATCH 19/28] tricore: replace cpu_tricore_init() " Igor Mammedov
2017-08-16 19:56   ` Eduardo Habkost
2017-07-14 13:52 ` [Qemu-devel] [PATCH 20/28] sh4: replace cpu_sh4_init() " Igor Mammedov
2017-08-16 19:57   ` Eduardo Habkost
2017-07-14 13:52 ` [Qemu-devel] [PATCH 21/28] arm: replace cpu_arm_init() " Igor Mammedov
2017-08-14  8:53   ` Andrew Jones
2017-07-14 13:52 ` [Qemu-devel] [PATCH 22/28] cris: replace cpu_cris_init() " Igor Mammedov
2017-08-16 19:57   ` Eduardo Habkost
2017-07-14 13:52 ` [Qemu-devel] [PATCH 23/28] x86: replace cpu_x86_init() " Igor Mammedov
2017-08-16 19:57   ` Eduardo Habkost
2017-07-14 13:52 ` [Qemu-devel] [PATCH 24/28] lm32: replace cpu_lm32_init() " Igor Mammedov
2017-07-14 15:51   ` Michael Walle
2017-08-16 19:58   ` Eduardo Habkost
2017-07-14 13:52 ` [Qemu-devel] [PATCH 25/28] moxie: replace cpu_moxie_init() " Igor Mammedov
2017-08-16 19:58   ` Eduardo Habkost
2017-07-14 13:52 ` [Qemu-devel] [PATCH 26/28] openrisc: replace cpu_openrisc_init() " Igor Mammedov
2017-08-16 19:58   ` Eduardo Habkost
2017-08-16 21:28   ` Stafford Horne
2017-07-14 13:52 ` [Qemu-devel] [PATCH 27/28] unicore32: replace uc32_cpu_init() " Igor Mammedov
2017-08-16 19:59   ` Eduardo Habkost
2017-07-14 13:52 ` [Qemu-devel] [PATCH 28/28] ppc: replace cpu_ppc_init() " Igor Mammedov
2017-07-15  2:36   ` David Gibson
2017-07-15  6:09   ` Hervé Poussineau
2017-08-16 19:59   ` Eduardo Habkost

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=20170821103241.2651a5bc@nial.brq.redhat.com \
    --to=imammedo@redhat.com \
    --cc=ehabkost@redhat.com \
    --cc=qemu-devel@nongnu.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.