All of lore.kernel.org
 help / color / mirror / Atom feed
From: Igor Mammedov <imammedo@redhat.com>
To: qemu-devel@nongnu.org
Cc: imammedo@redhat.com, mst@redhat.com,
	Ani Sinha <anisinha@redhat.com>,
	Fiona Ebner <f.ebner@proxmox.com>
Subject: [PATCH v3 16/20] smbios: in case of entry point is 'auto' try to build v2 tables 1st
Date: Tue, 12 Mar 2024 17:10:46 +0100	[thread overview]
Message-ID: <20240312161050.2248814-17-imammedo@redhat.com> (raw)
In-Reply-To: <20240312161050.2248814-1-imammedo@redhat.com>

QEMU for some time now uses SMBIOS 3.0 for PC/Q35 machines by
default, however Windows has a bug in locating SMBIOS 3.0
entrypoint and fails to find tables when booted on SeaBIOS
(on UEFI SMBIOS 3.0 tables work fine since firmware hands
over tables in another way)

Missing SMBIOS tables may lead to some issues for guest
though (worst are: possible reactiveation, inability to
get virtio drivers from 'Windows Update')

It's unclear  at this point if MS will fix the issue on their
side. So instead of it (or rather in addition) this patch
will try to workaround the issue.

aka, use smbios-entry-point-type=auto to make QEMU try
generating conservative SMBIOS 2.0 tables and if that
fails (due to limits/requested configuration) fallback
to SMBIOS 3.0 tables.

With this in place majority of users will use SMBIOS 2.0
tables which work fine with (Windows + legacy BIOS).
The configurations that is not to possible to describe
with SMBIOS 2.0 will switch automatically to SMBIOS 3.0
(which will trigger Windows bug but there is nothing
QEMU can do here, so go and aks Microsoft to real fix).

Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Reviewed-by: Ani Sinha <anisinha@redhat.com>
Tested-by: Fiona Ebner <f.ebner@proxmox.com>
---
 hw/smbios/smbios.c | 52 +++++++++++++++++++++++++++++++++++++++++++---
 1 file changed, 49 insertions(+), 3 deletions(-)

diff --git a/hw/smbios/smbios.c b/hw/smbios/smbios.c
index d3d89b4dd0..c4a953bbc9 100644
--- a/hw/smbios/smbios.c
+++ b/hw/smbios/smbios.c
@@ -959,7 +959,7 @@ static void smbios_entry_point_setup(SmbiosEntryPointType ep_type)
     }
 }
 
-void smbios_get_tables(MachineState *ms,
+static bool smbios_get_tables_ep(MachineState *ms,
                        SmbiosEntryPointType ep_type,
                        const struct smbios_phys_mem_area *mem_array,
                        const unsigned int mem_array_size,
@@ -968,6 +968,7 @@ void smbios_get_tables(MachineState *ms,
                        Error **errp)
 {
     unsigned i, dimm_cnt, offset;
+    ERRP_GUARD();
 
     assert(ep_type == SMBIOS_ENTRY_POINT_TYPE_32 ||
            ep_type == SMBIOS_ENTRY_POINT_TYPE_64);
@@ -1053,11 +1054,56 @@ void smbios_get_tables(MachineState *ms,
         abort();
     }
 
-    return;
+    return true;
 err_exit:
     g_free(smbios_tables);
     smbios_tables = NULL;
-    return;
+    return false;
+}
+
+void smbios_get_tables(MachineState *ms,
+                       SmbiosEntryPointType ep_type,
+                       const struct smbios_phys_mem_area *mem_array,
+                       const unsigned int mem_array_size,
+                       uint8_t **tables, size_t *tables_len,
+                       uint8_t **anchor, size_t *anchor_len,
+                       Error **errp)
+{
+    Error *local_err = NULL;
+    bool is_valid;
+    ERRP_GUARD();
+
+    switch (ep_type) {
+    case SMBIOS_ENTRY_POINT_TYPE_AUTO:
+    case SMBIOS_ENTRY_POINT_TYPE_32:
+        is_valid = smbios_get_tables_ep(ms, SMBIOS_ENTRY_POINT_TYPE_32,
+                                        mem_array, mem_array_size,
+                                        tables, tables_len,
+                                        anchor, anchor_len,
+                                        &local_err);
+        if (is_valid || ep_type != SMBIOS_ENTRY_POINT_TYPE_AUTO) {
+            break;
+        }
+        /*
+         * fall through in case AUTO endpoint is selected and
+         * SMBIOS 2.x tables can't be generated, to try if SMBIOS 3.x
+         * tables would work
+         */
+    case SMBIOS_ENTRY_POINT_TYPE_64:
+        error_free(local_err);
+        local_err = NULL;
+        is_valid = smbios_get_tables_ep(ms, SMBIOS_ENTRY_POINT_TYPE_64,
+                                        mem_array, mem_array_size,
+                                        tables, tables_len,
+                                        anchor, anchor_len,
+                                        &local_err);
+        break;
+    default:
+        abort();
+    }
+    if (!is_valid) {
+        error_propagate(errp, local_err);
+    }
 }
 
 static void save_opt(const char **dest, QemuOpts *opts, const char *name)
-- 
2.39.3



  parent reply	other threads:[~2024-03-12 16:14 UTC|newest]

Thread overview: 26+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-03-12 16:10 [PATCH v3 00/20] Workaround Windows failing to find 64bit SMBIOS entry point with SeaBIOS Igor Mammedov
2024-03-12 16:10 ` [PATCH v3 01/20] tests: smbios: make it possible to write SMBIOS only test Igor Mammedov
2024-03-12 16:10 ` [PATCH v3 02/20] tests: smbios: add test for -smbios type=11 option Igor Mammedov
2024-03-12 16:10 ` [PATCH v3 03/20] tests: smbios: add test for legacy mode CLI options Igor Mammedov
2024-03-12 16:10 ` [PATCH v3 04/20] smbios: cleanup smbios_get_tables() from legacy handling Igor Mammedov
2024-03-12 16:10 ` [PATCH v3 05/20] smbios: get rid of smbios_smp_sockets global Igor Mammedov
2024-03-12 16:10 ` [PATCH v3 06/20] smbios: get rid of smbios_legacy global Igor Mammedov
2024-03-12 16:10 ` [PATCH v3 07/20] smbios: avoid mangling user provided tables Igor Mammedov
2024-03-12 16:10 ` [PATCH v3 08/20] smbios: don't check type4 structures in legacy mode Igor Mammedov
2024-03-12 16:10 ` [PATCH v3 09/20] smbios: add smbios_add_usr_blob_size() helper Igor Mammedov
2024-03-12 16:10 ` [PATCH v3 10/20] smbios: rename/expose structures/bitmaps used by both legacy and modern code Igor Mammedov
2024-03-12 16:10 ` [PATCH v3 11/20] smbios: build legacy mode code only for 'pc' machine Igor Mammedov
2024-03-12 16:10 ` [PATCH v3 12/20] smbios: handle errors consistently Igor Mammedov
2024-03-12 16:10 ` [PATCH v3 13/20] smbios: get rid of global smbios_ep_type Igor Mammedov
2024-03-12 16:10 ` [PATCH v3 14/20] smbios: clear smbios_type4_count before building tables Igor Mammedov
2024-03-12 16:10 ` [PATCH v3 15/20] smbios: extend smbios-entry-point-type with 'auto' value Igor Mammedov
2024-03-12 16:10 ` Igor Mammedov [this message]
2024-03-12 16:10 ` [PATCH v3 17/20] smbios: error out when building type 4 table is not possible Igor Mammedov
2024-03-12 16:10 ` [PATCH v3 18/20] tests: acpi/smbios: whitelist expected blobs Igor Mammedov
2024-03-12 16:10 ` [PATCH v3 19/20] pc/q35: set SMBIOS entry point type to 'auto' by default Igor Mammedov
2024-03-12 16:10 ` [PATCH v3 20/20] tests: acpi: update expected SSDT.dimmpxm blob Igor Mammedov
2024-03-12 17:31 ` [PATCH v3 00/20] Workaround Windows failing to find 64bit SMBIOS entry point with SeaBIOS Michael S. Tsirkin
2024-03-13  8:49   ` Igor Mammedov
2024-03-13  9:10     ` Michael S. Tsirkin
2024-03-13 22:51     ` Michael S. Tsirkin
2024-03-12 17:37 ` Michael S. Tsirkin

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=20240312161050.2248814-17-imammedo@redhat.com \
    --to=imammedo@redhat.com \
    --cc=anisinha@redhat.com \
    --cc=f.ebner@proxmox.com \
    --cc=mst@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.