Received-SPF: pass (google.com: domain of linux-kernel+bounces-186670-linux.lists.archive=gmail.com@vger.kernel.org designates 2604:1380:45e3:2400::1 as permitted sender) client-ip=2604:1380:45e3:2400::1;
Message-ID: <d4496b4ed8a8a7bb34cf12e4cce65a6ad6705bc0.camel@redhat.com>
Subject: Re: Early boot regression from f0551af0213 ("x86/topology: Ignore
 non-present APIC IDs in a present package")
From: Lyude Paul <lyude@redhat.com>
To: Thomas Gleixner <tglx@linutronix.de>, "Linux regression tracking
 (Thorsten Leemhuis)" <regressions@leemhuis.info>
Cc: x86@kernel.org, linux-kernel@vger.kernel.org, Mario Limonciello
	 <mario.limonciello@amd.com>, Borislav Petkov <bp@alien8.de>, Linux kernel
 regressions list <regressions@lists.linux.dev>
Date: Wed, 22 May 2024 15:35:21 -0400
In-Reply-To: <87wmntkhak.ffs@tglx>
References: <3d77cb89857ee43a9c31249f4eab7196013bc4b4.camel@redhat.com>
	 <20240418082703.GCZiDZVyra7qOQbyqn@fat_crate.local>
	 <fd040809d95b3e12b2fdc78a2409e187716bc66f.camel@redhat.com>
	 <87plumxz4x.ffs@tglx>
	 <abbb7d7ca781f6c664e4c5b1dffc19394ac79691.camel@redhat.com>
	 <87le59vw1y.ffs@tglx>
	 <3a0afe545747e5314a9cb6bbaa9ce90b259ddfac.camel@redhat.com>
	 <87edautcmz.ffs@tglx>
	 <3b1d16e357c1f9badeef405366492f05af26c085.camel@redhat.com>
	 <878r11t8zu.ffs@tglx> <016902d9-3858-4c65-b3ec-f7a5103af63c@amd.com>
	 <51d0dff8-2888-463c-95ab-71b491f12a8f@leemhuis.info> <877cg4ppd5.ffs@tglx>
	 <ea927dad269cc21de1d0baf3d6c9f66ee025b862.camel@redhat.com>
	 <d2c6f335a6eb5892b0d894d5df4a6e713fa013b5.camel@redhat.com>
	 <87jzjxn6s5.ffs@tglx>
	 <d3fe5278e7cd5af6c62b470b281b547b67e3959a.camel@redhat.com>
	 <97bd95480a8b9951edc9ee2d2648d1b9c574e3b0.camel@redhat.com>
	 <87bk58n6le.ffs@tglx>
	 <2fd6009d21d606d13f0c472dbaa754a21f3105d9.camel@redhat.com>
	 <87wmntkhak.ffs@tglx>
Organization: Red Hat Inc.
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
User-Agent: Evolution 3.52.1 (3.52.1-1.fc40) 
Precedence: bulk
MIME-Version: 1.0

Awesome! This patch does seem to make the system boot, thank you for your h=
elp
<3

Tested-by: Lyude Paul <lyude@redhat.com>

On Thu, 2024-05-16 at 15:38 +0200, Thomas Gleixner wrote:
> Lyude!
>=20
> On Wed, May 15 2024 at 19:15, Lyude Paul wrote:
> > On Tue, 2024-05-14 at 10:25 +0200, Thomas Gleixner wrote:
> > >=20
> > > Which one of the debug patches did you use?
> >=20
> > The one you sent on 4/18, when you also asked me for the output of
> > /sys/kernel/debug/x86/topo/
> >=20
> > (I thought I remembered that patch not booting on previous kernels but =
maybe
> > I'm misremembering)
>=20
> It booted with the -rc kernel too. :)
>=20
> I found an interesting and probably related difference in the boot logs
> though. Up to 8 possible CPUs the kernel uses logical destination mode
> for the APIC. With more than 8 it uses physical destination mode.
>=20
> Can you please test the patch below on top of 6.9 and validate that it
> boots w/o any magic command line parameter?
>=20
> Thanks,
>=20
>         tglx
> ---
> diff --git a/arch/x86/kernel/apic/apic.c b/arch/x86/kernel/apic/apic.c
> index 66fd4b2a37a3..db5e93a7c194 100644
> --- a/arch/x86/kernel/apic/apic.c
> +++ b/arch/x86/kernel/apic/apic.c
> @@ -1883,6 +1883,8 @@ static inline void try_to_enable_x2apic(int remap_m=
ode) { }
>  static inline void __x2apic_enable(void) { }
>  #endif /* !CONFIG_X86_X2APIC */
> =20
> +int irq_remap_mode =3D -1;
> +
>  void __init enable_IR_x2apic(void)
>  {
>  	unsigned long flags;
> @@ -1915,6 +1917,8 @@ void __init enable_IR_x2apic(void)
> =20
>  	if (ir_stat < 0)
>  		restore_ioapic_entries();
> +	else
> +		irq_remap_mode =3D ir_stat;
>  	legacy_pic->restore_mask();
>  	local_irq_restore(flags);
>  }
> diff --git a/arch/x86/kernel/apic/apic_flat_64.c b/arch/x86/kernel/apic/a=
pic_flat_64.c
> index f37ad3392fec..5cc386db3557 100644
> --- a/arch/x86/kernel/apic/apic_flat_64.c
> +++ b/arch/x86/kernel/apic/apic_flat_64.c
> @@ -12,6 +12,7 @@
>  #include <linux/export.h>
>  #include <linux/acpi.h>
> =20
> +#include <asm/irq_remapping.h>
>  #include <asm/jailhouse_para.h>
>  #include <asm/apic.h>
> =20
> @@ -130,7 +131,8 @@ static int physflat_acpi_madt_oem_check(char *oem_id,=
 char *oem_table_id)
> =20
>  static int physflat_probe(void)
>  {
> -	return apic =3D=3D &apic_physflat || num_possible_cpus() > 8 || jailhou=
se_paravirt();
> +	return apic =3D=3D &apic_physflat || irq_remap_mode =3D=3D IRQ_REMAP_XA=
PIC_MODE ||
> +		num_possible_cpus() > 8 || jailhouse_paravirt();
>  }
> =20
>  static struct apic apic_physflat __ro_after_init =3D {
> diff --git a/arch/x86/kernel/apic/local.h b/arch/x86/kernel/apic/local.h
> index 842fe28496be..f633ab6dfa9f 100644
> --- a/arch/x86/kernel/apic/local.h
> +++ b/arch/x86/kernel/apic/local.h
> @@ -22,6 +22,8 @@ void x2apic_send_IPI_allbutself(int vector);
>  void x2apic_send_IPI_self(int vector);
>  extern u32 x2apic_max_apicid;
> =20
> +extern int irq_remap_mode;
> +
>  /* IPI */
> =20
>  DECLARE_STATIC_KEY_FALSE(apic_use_ipi_shorthand);
>=20

--=20
Cheers,
 Lyude Paul (she/her)
 Software Engineer at Red Hat