xen-devel.lists.xenproject.org archive mirror
 help / color / mirror / Atom feed
From: Wei Chen <Wei.Chen@arm.com>
To: Julien Grall <julien@xen.org>,
	"xen-devel@lists.xenproject.org" <xen-devel@lists.xenproject.org>,
	"sstabellini@kernel.org" <sstabellini@kernel.org>,
	"jbeulich@suse.com" <jbeulich@suse.com>
Cc: Bertrand Marquis <Bertrand.Marquis@arm.com>
Subject: RE: [XEN RFC PATCH 18/40] xen/arm: Keep memory nodes in dtb for NUMA when boot from EFI
Date: Fri, 20 Aug 2021 02:18:18 +0000	[thread overview]
Message-ID: <DB9PR08MB6857F9C9DAE5314C1F99E0159EC19@DB9PR08MB6857.eurprd08.prod.outlook.com> (raw)
In-Reply-To: <bfcfa593-60ac-43a7-4f1f-a7a3ab04a1ff@xen.org>

Hi Julien,

> -----Original Message-----
> From: Julien Grall <julien@xen.org>
> Sent: 2021年8月20日 1:35
> To: Wei Chen <Wei.Chen@arm.com>; xen-devel@lists.xenproject.org;
> sstabellini@kernel.org; jbeulich@suse.com
> Cc: Bertrand Marquis <Bertrand.Marquis@arm.com>
> Subject: Re: [XEN RFC PATCH 18/40] xen/arm: Keep memory nodes in dtb for
> NUMA when boot from EFI
> 
> Hi Wei,
> 
> On 11/08/2021 11:24, Wei Chen wrote:
> > EFI can get memory map from EFI system table. But EFI system
> > table doesn't contain memory NUMA information, EFI depends on
> > ACPI SRAT or device tree memory node to parse memory blocks'
> > NUMA mapping.
> >
> > But in current code, when Xen is booting from EFI, it will
> > delete all memory nodes in device tree. So in UEFI + DTB
> > boot, we don't have numa-node-id for memory blocks any more.
> >
> > So in this patch, we will keep memory nodes in device tree for
> > NUMA code to parse memory numa-node-id later.
> >
> > As a side effect, if we still parse boot memory information in
> > early_scan_node, bootmem.info will calculate memory ranges in
> > memory nodes twice. So we have to prvent early_scan_node to
> 
> s/prvent/prevent/
> 

Ok

> > parse memory nodes in EFI boot.
> >
> > As EFI APIs only can be used in Arm64, so we introduced a wrapper
> > in header file to prevent #ifdef CONFIG_ARM_64/32 in code block.
> >
> > Signed-off-by: Wei Chen <wei.chen@arm.com>
> > ---
> >   xen/arch/arm/bootfdt.c      |  8 +++++++-
> >   xen/arch/arm/efi/efi-boot.h | 25 -------------------------
> >   xen/include/asm-arm/setup.h |  6 ++++++
> >   3 files changed, 13 insertions(+), 26 deletions(-)
> >
> > diff --git a/xen/arch/arm/bootfdt.c b/xen/arch/arm/bootfdt.c
> > index 476e32e0f5..7df149dbca 100644
> > --- a/xen/arch/arm/bootfdt.c
> > +++ b/xen/arch/arm/bootfdt.c
> > @@ -11,6 +11,7 @@
> >   #include <xen/lib.h>
> >   #include <xen/kernel.h>
> >   #include <xen/init.h>
> > +#include <xen/efi.h>
> >   #include <xen/device_tree.h>
> >   #include <xen/libfdt/libfdt.h>
> >   #include <xen/sort.h>
> > @@ -335,7 +336,12 @@ static int __init early_scan_node(const void *fdt,
> >   {
> >       int rc = 0;
> >
> > -    if ( device_tree_node_matches(fdt, node, "memory") )
> > +    /*
> > +     * If system boot from EFI, bootinfo.mem has been set by EFI,
> 
> "If the system boot". Although, I would suggest to write:
> 
> "If Xen has been booted via UEFI, the memory banks will already be
> populated. So we should skip the parsing."
> 

Yes, that would be better. I will change it in next version.

> > +     * so we don't need to parse memory node from DTB.
> > +     */
> > +    if ( device_tree_node_matches(fdt, node, "memory") &&
> > +         !arch_efi_enabled(EFI_BOOT) )
> 
> arch_efi_enabled() is going to be less expensive than
> device_tree_node_matches(). So I would suggest to re-order the operands.
> 

yes.

> >           rc = process_memory_node(fdt, node, name, depth,
> >                                    address_cells, size_cells,
> &bootinfo.mem);
> >       else if ( depth == 1 && !dt_node_cmp(name, "reserved-memory") )
> > diff --git a/xen/arch/arm/efi/efi-boot.h b/xen/arch/arm/efi/efi-boot.h
> > index cf9c37153f..d0a9987fa4 100644
> > --- a/xen/arch/arm/efi/efi-boot.h
> > +++ b/xen/arch/arm/efi/efi-boot.h
> > @@ -197,33 +197,8 @@ EFI_STATUS __init
> fdt_add_uefi_nodes(EFI_SYSTEM_TABLE *sys_table,
> >       int status;
> >       u32 fdt_val32;
> >       u64 fdt_val64;
> > -    int prev;
> >       int num_rsv;
> >
> > -    /*
> > -     * Delete any memory nodes present.  The EFI memory map is the only
> > -     * memory description provided to Xen.
> > -     */
> > -    prev = 0;
> > -    for (;;)
> > -    {
> > -        const char *type;
> > -        int len;
> > -
> > -        node = fdt_next_node(fdt, prev, NULL);
> > -        if ( node < 0 )
> > -            break;
> > -
> > -        type = fdt_getprop(fdt, node, "device_type", &len);
> > -        if ( type && strncmp(type, "memory", len) == 0 )
> > -        {
> > -            fdt_del_node(fdt, node);
> > -            continue;
> > -        }
> > -
> > -        prev = node;
> > -    }
> > -
> >      /*
> >       * Delete all memory reserve map entries. When booting via UEFI,
> >       * kernel will use the UEFI memory map to find reserved regions.
> > diff --git a/xen/include/asm-arm/setup.h b/xen/include/asm-arm/setup.h
> > index c4b6af6029..e4fb5f0d49 100644
> > --- a/xen/include/asm-arm/setup.h
> > +++ b/xen/include/asm-arm/setup.h
> > @@ -123,6 +123,12 @@ void device_tree_get_reg(const __be32 **cell, u32
> address_cells,
> >   u32 device_tree_get_u32(const void *fdt, int node,
> >                           const char *prop_name, u32 dflt);
> >
> > +#if defined(CONFIG_ARM_64)
> > +#define arch_efi_enabled(x) efi_enabled(x)
> > +#else
> > +#define arch_efi_enabled(x) (0)
> > +#endif
> 
> I would prefer if we introduce CONFIG_EFI that would stub efi_enabled
> for architecture not supporting EFI.
> 

Yes, that's a good idea.
I also feel a little awkward with the current arch helpers.

> Cheers,
> 
> --
> Julien Grall

  reply	other threads:[~2021-08-20  2:19 UTC|newest]

Thread overview: 196+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-08-11 10:23 [XEN RFC PATCH 00/40] Add device tree based NUMA support to Arm64 Wei Chen
2021-08-11 10:23 ` [XEN RFC PATCH 01/40] tools: Fix -Werror=maybe-uninitialized for xlu_pci_parse_bdf Wei Chen
2021-08-11 10:49   ` Jan Beulich
2021-08-13  6:28     ` Wei Chen
2021-08-11 10:23 ` [XEN RFC PATCH 02/40] xen/arm: Print a 64-bit number in hex from early uart Wei Chen
2021-08-19 13:05   ` Julien Grall
2021-08-20  1:13     ` Wei Chen
2021-08-11 10:23 ` [XEN RFC PATCH 03/40] xen/x86: Initialize memnodemapsize while faking NUMA node Wei Chen
2021-08-12 15:32   ` Jan Beulich
2021-08-13  7:26     ` Wei Chen
2021-08-13  8:29       ` Jan Beulich
2021-08-11 10:23 ` [XEN RFC PATCH 04/40] xen/arm: return default DMA bit width when platform is not set Wei Chen
2021-08-11 10:54   ` Jan Beulich
2021-08-13  6:54     ` Wei Chen
2021-08-13  6:56       ` Jan Beulich
2021-08-19 13:28   ` Julien Grall
2021-08-20  2:04     ` Wei Chen
2021-08-20  8:20       ` Julien Grall
2021-08-20  9:37         ` Wei Chen
2021-08-20 11:18           ` Julien Grall
2021-08-20 11:58             ` Wei Chen
2021-08-11 10:23 ` [XEN RFC PATCH 05/40] xen/arm: Fix lowmem_bitsize when arch_get_dma_bitsize return 0 Wei Chen
2021-08-19 13:32   ` Julien Grall
2021-08-20  2:05     ` Wei Chen
2021-08-11 10:23 ` [XEN RFC PATCH 06/40] xen: decouple NUMA from ACPI in Kconfig Wei Chen
2021-08-12 15:36   ` Jan Beulich
2021-08-13  7:27     ` Wei Chen
2021-08-12 16:54   ` Julien Grall
2021-08-13  7:28     ` Wei Chen
2021-08-11 10:23 ` [XEN RFC PATCH 07/40] xen/arm: use !CONFIG_NUMA to keep fake NUMA API Wei Chen
2021-08-19 13:34   ` Julien Grall
2021-08-20  2:08     ` Wei Chen
2021-08-20  8:23       ` Julien Grall
2021-08-20 10:24         ` Wei Chen
2021-08-20 11:24           ` Julien Grall
2021-08-20 12:23             ` Wei Chen
2021-08-20 14:41               ` Julien Grall
2021-08-11 10:23 ` [XEN RFC PATCH 08/40] xen/x86: Move NUMA memory node map functions to common Wei Chen
2021-08-23 17:47   ` Julien Grall
2021-08-24  4:07     ` Wei Chen
2021-08-11 10:23 ` [XEN RFC PATCH 09/40] xen/x86: Move numa_add_cpu_node " Wei Chen
2021-08-23 17:54   ` Julien Grall
2021-08-24  4:18     ` Wei Chen
2021-08-11 10:23 ` [XEN RFC PATCH 10/40] xen/x86: Move NR_NODE_MEMBLKS macro " Wei Chen
2021-08-23 17:58   ` Julien Grall
2021-08-11 10:23 ` [XEN RFC PATCH 11/40] xen/x86: Move NUMA nodes and memory block ranges " Wei Chen
2021-08-24 17:40   ` Julien Grall
2021-08-25  0:57     ` Wei Chen
2021-08-11 10:23 ` [XEN RFC PATCH 12/40] xen/x86: Move numa_initmem_init " Wei Chen
2021-08-25 10:21   ` Julien Grall
2021-08-25 11:15     ` Wei Chen
2021-08-25 13:26       ` Julien Grall
2021-08-11 10:23 ` [XEN RFC PATCH 13/40] xen/arm: introduce numa_set_node for Arm Wei Chen
2021-08-25 10:36   ` Julien Grall
2021-08-25 12:07     ` Wei Chen
2021-08-25 13:24       ` Julien Grall
2021-08-26  5:13         ` Wei Chen
2021-08-11 10:23 ` [XEN RFC PATCH 14/40] xen/arm: set NUMA nodes max number to 64 by default Wei Chen
2021-08-25 13:28   ` Julien Grall
2021-08-25 13:36     ` Jan Beulich
2021-08-26  2:26       ` Wei Chen
2021-08-11 10:23 ` [XEN RFC PATCH 15/40] xen/x86: move NUMA API from x86 header to common header Wei Chen
2021-08-11 10:23 ` [XEN RFC PATCH 16/40] xen/arm: Create a fake NUMA node to use common code Wei Chen
2021-08-26 23:10   ` Stefano Stabellini
2021-08-27  1:15     ` Wei Chen
2021-08-27  6:18     ` Jan Beulich
2021-08-27  9:32       ` Wei Chen
2021-08-11 10:24 ` [XEN RFC PATCH 17/40] xen/arm: Introduce DEVICE_TREE_NUMA Kconfig for arm64 Wei Chen
2021-08-19 13:38   ` Julien Grall
2021-08-20  2:30     ` Wei Chen
2021-08-20  8:41       ` Julien Grall
2021-08-20 10:49         ` Wei Chen
2021-08-20 11:28           ` Julien Grall
2021-08-20 12:25             ` Wei Chen
2021-08-11 10:24 ` [XEN RFC PATCH 18/40] xen/arm: Keep memory nodes in dtb for NUMA when boot from EFI Wei Chen
2021-08-19 17:35   ` Julien Grall
2021-08-20  2:18     ` Wei Chen [this message]
2021-08-26 23:24   ` Stefano Stabellini
2021-08-27  7:41     ` Julien Grall
2021-08-27 23:10       ` Stefano Stabellini
2021-08-27  9:23     ` Wei Chen
2021-08-11 10:24 ` [XEN RFC PATCH 19/40] xen: fdt: Introduce a helper to check fdt node type Wei Chen
2021-08-25 13:39   ` Julien Grall
2021-08-26  6:00     ` Wei Chen
2021-08-11 10:24 ` [XEN RFC PATCH 20/40] xen/arm: implement node distance helpers for Arm64 Wei Chen
2021-08-26 23:52   ` Stefano Stabellini
2021-08-27  9:30     ` Wei Chen
2021-08-11 10:24 ` [XEN RFC PATCH 21/40] xen/arm: introduce device_tree_numa as a switch for device tree NUMA Wei Chen
2021-08-19 17:45   ` Julien Grall
2021-08-20  2:21     ` Wei Chen
2021-08-11 10:24 ` [XEN RFC PATCH 22/40] xen/arm: introduce a helper to parse device tree processor node Wei Chen
2021-08-19 18:09   ` Julien Grall
2021-08-23  8:42     ` Wei Chen
2021-08-19 18:10   ` Julien Grall
2021-08-23  8:47     ` Wei Chen
2021-08-23 10:59       ` Julien Grall
2021-08-24  4:09         ` Wei Chen
2021-08-19 18:13   ` Julien Grall
2021-08-20  2:23     ` Wei Chen
2021-08-20  8:44       ` Julien Grall
2021-08-20 11:53         ` Wei Chen
2021-08-27  0:06   ` Stefano Stabellini
2021-08-27  9:31     ` Wei Chen
2021-08-11 10:24 ` [XEN RFC PATCH 23/40] xen/arm: introduce a helper to parse device tree memory node Wei Chen
2021-08-25 13:48   ` Julien Grall
2021-08-26  6:35     ` Wei Chen
2021-08-26  8:21       ` Julien Grall
2021-08-26 11:54         ` Wei Chen
2021-08-28  1:06   ` Stefano Stabellini
2021-08-28  3:56     ` Wei Chen
2021-08-28 10:33       ` Julien Grall
2021-08-28 13:58         ` Wei Chen
2021-09-08  7:34           ` Wei Chen
2021-09-08 22:31             ` Stefano Stabellini
2021-09-09  3:54               ` Wei Chen
2021-08-11 10:24 ` [XEN RFC PATCH 24/40] xen/arm: introduce a helper to parse device tree NUMA distance map Wei Chen
2021-08-25 13:56   ` Julien Grall
2021-08-26  7:01     ` Wei Chen
2021-08-31  0:48   ` Stefano Stabellini
2021-08-31 10:17     ` Wei Chen
2021-08-31 21:36       ` Stefano Stabellini
2021-09-01 11:04         ` Wei Chen
2021-09-01 16:21           ` Stefano Stabellini
2021-09-02  2:30             ` Wei Chen
2021-09-02 15:19               ` Stefano Stabellini
2021-09-02  6:00             ` Jan Beulich
2021-09-02 14:14               ` Wei Chen
2021-08-11 10:24 ` [XEN RFC PATCH 25/40] xen/arm: unified entry to parse all NUMA data from device tree Wei Chen
2021-08-31  0:54   ` Stefano Stabellini
2021-08-31 17:47     ` Julien Grall
2021-09-01 18:30       ` Stefano Stabellini
2021-09-02  2:48         ` Wei Chen
2021-08-11 10:24 ` [XEN RFC PATCH 26/40] xen/arm: Add boot and secondary CPU to NUMA system Wei Chen
2021-08-25 16:58   ` Julien Grall
2021-08-26  7:24     ` Wei Chen
2021-08-26  8:49       ` Julien Grall
2021-08-26  9:39         ` Jan Beulich
2021-08-26 12:08           ` Wei Chen
2021-08-26 12:26             ` Jan Beulich
2021-08-11 10:24 ` [XEN RFC PATCH 27/40] xen/arm: build CPU NUMA node map while creating cpu_logical_map Wei Chen
2021-08-25 17:06   ` Julien Grall
2021-08-26  7:26     ` Wei Chen
2021-08-11 10:24 ` [XEN RFC PATCH 28/40] xen/x86: decouple nodes_cover_memory with E820 map Wei Chen
2021-08-31  1:07   ` Stefano Stabellini
2021-08-31 10:19     ` Wei Chen
2021-08-11 10:24 ` [XEN RFC PATCH 29/40] xen/arm: implement Arm arch helpers Arm to get memory map info Wei Chen
2021-08-25 17:09   ` Julien Grall
2021-08-26  7:27     ` Wei Chen
2021-08-11 10:24 ` [XEN RFC PATCH 30/40] xen: move NUMA memory and CPU parsed nodemasks to common Wei Chen
2021-08-25 17:16   ` Julien Grall
2021-08-26  7:29     ` Wei Chen
2021-08-11 10:24 ` [XEN RFC PATCH 31/40] xen/x86: move nodes_cover_memory " Wei Chen
2021-08-31  1:16   ` Stefano Stabellini
2021-08-31 13:43     ` Wei Chen
2021-08-11 10:24 ` [XEN RFC PATCH 32/40] xen/x86: make acpi_scan_nodes to be neutral Wei Chen
2021-08-27 14:08   ` Julien Grall
2021-08-28  2:11     ` Wei Chen
2021-08-11 10:24 ` [XEN RFC PATCH 33/40] xen: export bad_srat and srat_disabled to extern Wei Chen
2021-08-11 10:24 ` [XEN RFC PATCH 34/40] xen: move numa_scan_nodes from x86 to common Wei Chen
2021-08-27 14:14   ` Julien Grall
2021-08-28  2:12     ` Wei Chen
2021-08-31  1:26   ` Stefano Stabellini
2021-08-31 13:43     ` Wei Chen
2021-08-11 10:24 ` [XEN RFC PATCH 35/40] xen: enable numa_scan_nodes for device tree based NUMA Wei Chen
2021-08-27 14:19   ` Julien Grall
2021-08-28  2:13     ` Wei Chen
2021-08-11 10:24 ` [XEN RFC PATCH 36/40] xen/arm: keep guest still be NUMA unware Wei Chen
2021-08-27 14:28   ` Julien Grall
2021-08-28  2:19     ` Wei Chen
2021-08-11 10:24 ` [XEN RFC PATCH 37/40] xen: introduce an arch helper to do NUMA init failed fallback Wei Chen
2021-08-27 14:30   ` Julien Grall
2021-08-28  3:09     ` Wei Chen
2021-08-28  3:45       ` Wei Chen
2021-08-30  9:52         ` Jan Beulich
2021-08-30 10:38           ` Wei Chen
2021-08-11 10:24 ` [XEN RFC PATCH 38/40] xen/arm: enable device tree based NUMA in system init Wei Chen
2021-08-27 14:32   ` Julien Grall
2021-08-28  3:17     ` Wei Chen
2021-08-28 10:45       ` Julien Grall
2021-08-28 14:02         ` Wei Chen
2021-08-31  1:50   ` Stefano Stabellini
2021-08-31 13:43     ` Wei Chen
2021-08-11 10:24 ` [XEN RFC PATCH 39/40] xen/x86: move numa_setup to common to support NUMA switch in command line Wei Chen
2021-08-27 14:37   ` Julien Grall
2021-08-28  3:22     ` Wei Chen
2021-08-31  1:53   ` Stefano Stabellini
2021-08-31 13:44     ` Wei Chen
2021-08-11 10:24 ` [XEN RFC PATCH 40/40] xen/x86: move dump_numa info hotkey to common Wei Chen
2021-08-11 10:41 ` [XEN RFC PATCH 00/40] Add device tree based NUMA support to Arm64 Jan Beulich
2021-08-13  2:33   ` Wei Chen
2021-08-13  6:53     ` Jan Beulich
2021-08-19 13:42 ` Julien Grall
2021-08-19 14:05   ` Bertrand Marquis
2021-08-19 17:11     ` Julien Grall
2021-08-26  0:09 ` Stefano Stabellini
2021-08-26  7:31   ` Wei Chen

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=DB9PR08MB6857F9C9DAE5314C1F99E0159EC19@DB9PR08MB6857.eurprd08.prod.outlook.com \
    --to=wei.chen@arm.com \
    --cc=Bertrand.Marquis@arm.com \
    --cc=jbeulich@suse.com \
    --cc=julien@xen.org \
    --cc=sstabellini@kernel.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).