Skip to contentSkip to conversation

Support Secure Boot for multiboot2 Xen

PATCH In review Bobby Eshleman 4 participants 30 messages Jan 22, 2021 – Jun 9, 2021

lore
Share & tools
lore
#0 lore raw

v3 cover Support Secure Boot for multiboot2 Xen

This is version 3 for a patch set sent out to the ML in 2018 [1] to support UEFI Secure Boot for Xen on multiboot2 platforms.

A new binary, xen.mb.efi, is built. It contains the mb2 header as well as a hand-crafted PE/COFF header. The dom0 kernel is verified using the shim lock protocol.

I followed with v2 feedback and attempted to convert the PE/COFF header into C instead of ASM. Unfortunately, this was only possible for the first part (Legacy) of the PE/COFF header. The other parts required addresses only available at link time (such as __2M_rwdata_end, __pe_SizeOfImage, efi_mb_start address, etc...), which effectively ruled out C.

The biggest difference between v2 and v3 is that in v3 we do not attempt to merge xen.mb.efi and xen.efi into a single binary. Instead, this will be left to a future patch set, unless requested otherwise.

[1]: https://lists.xen.org/archives/html/xen-devel/2018-06/msg01292.html

Changes in v3:

- add requested comment clarification
- remove unnecessary fake data from PE/COFF head (like linker versions)
- macro-ize and refactor Makefile according to Jan's feedback
- break PE/COFF header into its own file
- shrink the PE/COFF to start 0x40 instead of 0x80 (my tests showed this function with no problem, on a live nested vm or using objdump/objcopy)
- support SOURCE_EPOCH for posix time
- removed date invocation that would break on FreeBSD
- style changes
- And obviously, ported to current HEAD

Daniel Kiper (5):
xen: add XEN_BUILD_POSIX_TIME
xen/x86: manually build xen.mb.efi binary xen/x86: add some addresses to the Multiboot header xen/x86: add some addresses to the Multiboot2 header xen/x86/efi: Verify dom0 kernel with SHIM_LOCK protocol in
efi_multiboot2()

 xen/Makefile                 |  22 ++++---
 xen/arch/x86/Makefile        |   7 +-
 xen/arch/x86/arch.mk         |   2 +
 xen/arch/x86/boot/Makefile   |   1 +
 xen/arch/x86/boot/head.S     |  53 +++++++++++++--
 xen/arch/x86/boot/pecoff.S   | 123 +++++++++++++++++++++++++++++++++++
 xen/arch/x86/efi/efi-boot.h  |  30 ++++++++-
 xen/arch/x86/efi/stub.c      |  17 ++++-
 xen/arch/x86/xen.lds.S       |  34 ++++++++++
 xen/common/efi/boot.c        |  19 ++++--
 xen/include/xen/compile.h.in |   1 +
 xen/include/xen/efi.h        |   1 +
 12 files changed, 283 insertions(+), 27 deletions(-)
 create mode 100644 xen/arch/x86/boot/pecoff.S
Signature · 1 line
2.30.0
#1 lore raw

PATCH v3 1/5 xen: add XEN_BUILD_POSIX_TIME

From: Daniel Kiper <[email protected]>

POSIX time is required to fill the TimeDateStamp field in the PE header.

Use SOURCE_DATE_EPOCH if available, otherwise use
date +%s (available on both Linux and FreeBSD).

---
 xen/Makefile                 | 2 ++
 xen/include/xen/compile.h.in | 1 +
 2 files changed, 3 insertions(+)
xen/Makefile +2 -0
index 544cc0995d..85f9b763a4 100644
@@ -11,6 +11,7 @@ export XEN_DOMAIN	?= $(shell ([ -x /bin/dnsdomainname ] && /bin/dnsdomainname) |
 export XEN_BUILD_DATE	?= $(shell LC_ALL=C date)
 export XEN_BUILD_TIME	?= $(shell LC_ALL=C date +%T)
 export XEN_BUILD_HOST	?= $(shell hostname)
+export XEN_BUILD_POSIX_TIME	?= $(shell echo $${SOURCE_DATE_EPOCH:-$(shell date +%s)})
 
 # Best effort attempt to find a python interpreter, defaulting to Python 3 if
 # available.  Fall back to just `python` if `which` is nowhere to be found.
@@ -386,6 +387,7 @@ delete-unfresh-files:
 include/xen/compile.h: include/xen/compile.h.in .banner
 	@sed -e 's/@@date@@/$(XEN_BUILD_DATE)/g' \
 	    -e 's/@@time@@/$(XEN_BUILD_TIME)/g' \
+	    -e 's/@@posix_time@@/$(XEN_BUILD_POSIX_TIME)/g' \
 	    -e 's/@@whoami@@/$(XEN_WHOAMI)/g' \
 	    -e 's/@@domain@@/$(XEN_DOMAIN)/g' \
 	    -e 's/@@hostname@@/$(XEN_BUILD_HOST)/g' \
xen/include/xen/compile.h.in +1 -0
index 440ecb25c1..b2ae6f96ad 100644
@@ -1,5 +1,6 @@
 #define XEN_COMPILE_DATE	"@@date@@"
 #define XEN_COMPILE_TIME	"@@time@@"
+#define XEN_COMPILE_POSIX_TIME	@@posix_time@@
 #define XEN_COMPILE_BY		"@@whoami@@"
 #define XEN_COMPILE_DOMAIN	"@@domain@@"
 #define XEN_COMPILE_HOST	"@@hostname@@"
Signature · 1 line
2.30.0
#2 lore raw

Re: [PATCH v3 1/5] xen: add XEN_BUILD_POSIX_TIME

Quoted 10 lines · Bobby Eshleman

From: Daniel Kiper <[email protected]>

POSIX time is required to fill the TimeDateStamp field in the PE header.

Use SOURCE_DATE_EPOCH if available, otherwise use
date +%s (available on both Linux and FreeBSD).

Especially with there not being any user of the new item, you will want to at least briefly explain ...

xen/Makefile export XEN_DOMAIN ?= $(shell ([ -x /bin/dnsdomainname ] && /bin/dnsdomainname) | · line 14

Quoted 7 lines
xen/Makefile +1 -0
@@ -11,6 +11,7 @@ export XEN_DOMAIN	?= $(shell ([ -x /bin/dnsdomainname ] && /bin/dnsdomainname) |
 export XEN_BUILD_DATE	?= $(shell LC_ALL=C date)
 export XEN_BUILD_TIME	?= $(shell LC_ALL=C date +%T)
 export XEN_BUILD_HOST	?= $(shell hostname)
+export XEN_BUILD_POSIX_TIME	?= $(shell echo $${SOURCE_DATE_EPOCH:-$(shell date +%s)})

... the use of SOURCE_DATE_EPOCH here when it's not used for XEN_BUILD_TIME (the two could also do with living side by side) and ...

xen/include/xen/compile.h.in line 6

Quoted 9 lines
xen/include/xen/compile.h.in +1 -0
@@ -1,5 +1,6 @@
 #define XEN_COMPILE_DATE	"@@date@@"
 #define XEN_COMPILE_TIME	"@@time@@"
+#define XEN_COMPILE_POSIX_TIME	@@posix_time@@
 #define XEN_COMPILE_BY		"@@whoami@@"
 #define XEN_COMPILE_DOMAIN	"@@domain@@"
 #define XEN_COMPILE_HOST	"@@hostname@@"

... the lack of quotes here when all neighboring items have them.

Jan

#3 lore raw

Re: [PATCH v3 1/5] xen: add XEN_BUILD_POSIX_TIME

Quoted 10 lines · Jan Beulich
Quoted 4 lines · Bobby Eshleman

export XEN_BUILD_DATE ?= $(shell LC_ALL=C date)
export XEN_BUILD_TIME ?= $(shell LC_ALL=C date +%T)
export XEN_BUILD_HOST ?= $(shell hostname)
+export XEN_BUILD_POSIX_TIME ?= $(shell echo $${SOURCE_DATE_EPOCH:-$(shell date +%s)})

... the use of SOURCE_DATE_EPOCH here when it's not used for XEN_BUILD_TIME (the two could also do with living side by side) and ...

XEN_BUILD_TIME is of the form "HH:MM:SS" and SOURCE_DATE_EPOCH / date +%s are unix timestamps (seconds since epoch). On Linux, date -d could be used to equalize the two timestamps... I'm not sure about FreeBSD, as -d is not required by POSIX.

I could place them side-by-side if that's preferred. I placed it afterwards here so that there wasn't one oddly aligned "?=" assignment in the middle of the others, as in rev2 it was requested their alignment be retained here.

Quoted 4 lines
Quoted 4 lines
xen/include/xen/compile.h.in +0 -0
@@ -1,5 +1,6 @@
 #define XEN_COMPILE_DATE	"@@date@@"
Quoted 9 lines

xen/include/xen/compile.h.in line 6

Quoted 5 lines
xen/include/xen/compile.h.in +1 -0
 #define XEN_COMPILE_TIME	"@@time@@"
+#define XEN_COMPILE_POSIX_TIME	@@posix_time@@
 #define XEN_COMPILE_BY		"@@whoami@@"
 #define XEN_COMPILE_DOMAIN	"@@domain@@"
 #define XEN_COMPILE_HOST	"@@hostname@@"

... the lack of quotes here when all neighboring items have them.

XEN_COMPILE_POSIX_TIME is used as a long, while the others are used as strings. Should this be commented?

Thank you for the feedback.

Best,
Bobby

#4 lore raw

Re: [PATCH v3 1/5] xen: add XEN_BUILD_POSIX_TIME

Quoted 17 lines · Bobby Eshleman
Quoted 10 lines · Jan Beulich
Quoted 4 lines · Bobby Eshleman

export XEN_BUILD_DATE ?= $(shell LC_ALL=C date)
export XEN_BUILD_TIME ?= $(shell LC_ALL=C date +%T)
export XEN_BUILD_HOST ?= $(shell hostname)
+export XEN_BUILD_POSIX_TIME ?= $(shell echo $${SOURCE_DATE_EPOCH:-$(shell date +%s)})

... the use of SOURCE_DATE_EPOCH here when it's not used for XEN_BUILD_TIME (the two could also do with living side by side) and ...

XEN_BUILD_TIME is of the form "HH:MM:SS" and SOURCE_DATE_EPOCH / date +%s are unix timestamps (seconds since epoch). On Linux, date -d could be used to equalize the two timestamps... I'm not sure about FreeBSD, as -d is not required by POSIX.

Quoted 4 lines

I could place them side-by-side if that's preferred. I placed it afterwards here so that there wasn't one oddly aligned "?=" assignment in the middle of the others, as in rev2 it was requested their alignment be retained here.

Personally I'd prefer if the time ones were all together. The odd alignment isn't uncommon when introducing new items with unexpectedly long names into a previously aligned list of items. Of course you will want to drop the excess blanks in any event - one each suffice as separators.

Quoted 14 lines
Quoted 4 lines
Quoted 4 lines
xen/include/xen/compile.h.in +0 -0
@@ -1,5 +1,6 @@
 #define XEN_COMPILE_DATE	"@@date@@"
Quoted 9 lines

xen/include/xen/compile.h.in line 6

Quoted 5 lines
xen/include/xen/compile.h.in +1 -0
 #define XEN_COMPILE_TIME	"@@time@@"
+#define XEN_COMPILE_POSIX_TIME	@@posix_time@@
 #define XEN_COMPILE_BY		"@@whoami@@"
 #define XEN_COMPILE_DOMAIN	"@@domain@@"
 #define XEN_COMPILE_HOST	"@@hostname@@"

... the lack of quotes here when all neighboring items have them.

Quoted 2 lines

XEN_COMPILE_POSIX_TIME is used as a long, while the others are used as strings. Should this be commented?

Not sure about commenting, but at the very least the description will imo want to point out the (intentional) difference. You shouldn't imply readers know in advance what a subsequent patch may use this for.

Jan

#5 lore raw

PATCH v3 2/5 xen/x86: manually build xen.mb.efi binary

In reply to Bobby Eshleman

From: Daniel Kiper <[email protected]>

This patch introduces xen.mb.efi which contains a manually built PE header.

This allows us to support Xen on UEFI Secure Boot-enabled hosts via multiboot2.

xen.mb.efi is a single binary that is loadable by a UEFI loader or via the Multiboot/Multiboot2 protocols.

---
 xen/Makefile                |  20 +++---
 xen/arch/x86/Makefile       |   7 +-
 xen/arch/x86/arch.mk        |   2 +
 xen/arch/x86/boot/Makefile  |   1 +
 xen/arch/x86/boot/head.S    |   1 +
 xen/arch/x86/boot/pecoff.S  | 123 ++++++++++++++++++++++++++++++++++++
 xen/arch/x86/efi/efi-boot.h |  24 ++++++-
 xen/arch/x86/efi/stub.c     |  12 +++-
 xen/arch/x86/xen.lds.S      |  34 ++++++++++
 xen/include/xen/efi.h       |   1 +
 10 files changed, 212 insertions(+), 13 deletions(-)
 create mode 100644 xen/arch/x86/boot/pecoff.S
xen/Makefile +11 -9
index 85f9b763a4..9c689c2095 100644
@@ -266,29 +266,31 @@ endif
 .PHONY: _build
 _build: $(TARGET)$(CONFIG_XEN_INSTALL_SUFFIX)
 
+define install_xen_links
+	$(INSTALL_DATA) $(TARGET)$1 $(D)$(BOOT_DIR)/$(T)-$(XEN_FULLVERSION)$1
+	ln -f -s $(T)-$(XEN_FULLVERSION)$1 $(D)$(BOOT_DIR)/$(T)-$(XEN_VERSION).$(XEN_SUBVERSION)$1
+	ln -f -s $(T)-$(XEN_FULLVERSION)$1 $(D)$(BOOT_DIR)/$(T)-$(XEN_VERSION)$1
+	ln -f -s $(T)-$(XEN_FULLVERSION)$1 $(D)$(BOOT_DIR)/$(T)$1
+endef
+
 .PHONY: _install
 _install: D=$(DESTDIR)
 _install: T=$(notdir $(TARGET))
 _install: Z=$(CONFIG_XEN_INSTALL_SUFFIX)
 _install: $(TARGET)$(CONFIG_XEN_INSTALL_SUFFIX)
 	[ -d $(D)$(BOOT_DIR) ] || $(INSTALL_DIR) $(D)$(BOOT_DIR)
-	$(INSTALL_DATA) $(TARGET)$(Z) $(D)$(BOOT_DIR)/$(T)-$(XEN_FULLVERSION)$(Z)
-	ln -f -s $(T)-$(XEN_FULLVERSION)$(Z) $(D)$(BOOT_DIR)/$(T)-$(XEN_VERSION).$(XEN_SUBVERSION)$(Z)
-	ln -f -s $(T)-$(XEN_FULLVERSION)$(Z) $(D)$(BOOT_DIR)/$(T)-$(XEN_VERSION)$(Z)
-	ln -f -s $(T)-$(XEN_FULLVERSION)$(Z) $(D)$(BOOT_DIR)/$(T)$(Z)
+	$(call install_xen_links,$(Z))
+	$(call install_xen_links,.mb.efi)
 	[ -d "$(D)$(DEBUG_DIR)" ] || $(INSTALL_DIR) $(D)$(DEBUG_DIR)
 	$(INSTALL_DATA) $(TARGET)-syms $(D)$(DEBUG_DIR)/$(T)-syms-$(XEN_FULLVERSION)
 	$(INSTALL_DATA) $(TARGET)-syms.map $(D)$(DEBUG_DIR)/$(T)-syms-$(XEN_FULLVERSION).map
 	$(INSTALL_DATA) $(KCONFIG_CONFIG) $(D)$(BOOT_DIR)/$(T)-$(XEN_FULLVERSION).config
 	if [ -r $(TARGET).efi -a -n '$(EFI_DIR)' ]; then \
 		[ -d $(D)$(EFI_DIR) ] || $(INSTALL_DIR) $(D)$(EFI_DIR); \
-		$(INSTALL_DATA) $(TARGET).efi $(D)$(EFI_DIR)/$(T)-$(XEN_FULLVERSION).efi; \
 		if [ -e $(TARGET).efi.map ]; then \
 			$(INSTALL_DATA) $(TARGET).efi.map $(D)$(DEBUG_DIR)/$(T)-$(XEN_FULLVERSION).efi.map; \
 		fi; \
-		ln -sf $(T)-$(XEN_FULLVERSION).efi $(D)$(EFI_DIR)/$(T)-$(XEN_VERSION).$(XEN_SUBVERSION).efi; \
-		ln -sf $(T)-$(XEN_FULLVERSION).efi $(D)$(EFI_DIR)/$(T)-$(XEN_VERSION).efi; \
-		ln -sf $(T)-$(XEN_FULLVERSION).efi $(D)$(EFI_DIR)/$(T).efi; \
+		$(call install_xen_links,.efi)) \
 		if [ -n '$(EFI_MOUNTPOINT)' -a -n '$(EFI_VENDOR)' ]; then \
 			$(INSTALL_DATA) $(TARGET).efi $(D)$(EFI_MOUNTPOINT)/efi/$(EFI_VENDOR)/$(T)-$(XEN_FULLVERSION).efi; \
 		elif [ "$(D)" = "$(patsubst $(shell cd $(XEN_ROOT) && pwd)/%,%,$(D))" ]; then \
@@ -341,7 +343,7 @@ _clean: delete-unfresh-files
 	$(MAKE) -f $(BASEDIR)/tools/kconfig/Makefile.kconfig ARCH=$(ARCH) SRCARCH=$(SRCARCH) clean
 	find . \( -name "*.o" -o -name ".*.d" -o -name ".*.d2" \
 		-o -name "*.gcno" -o -name ".*.cmd" \) -exec rm -f {} \;
-	rm -f include/asm $(TARGET) $(TARGET).gz $(TARGET).efi $(TARGET).efi.map $(TARGET)-syms $(TARGET)-syms.map *~ core
+	rm -f include/asm $(TARGET) $(TARGET).gz $(TARGET)*.efi $(TARGET).efi.map $(TARGET)-syms $(TARGET)-syms.map *~ core
 	rm -f include/asm-*/asm-offsets.h
 	rm -f .banner
 
xen/arch/x86/Makefile +6 -1
index 7769bb40d7..49c61adabf 100644
@@ -110,7 +110,7 @@ syms-warn-dup-$(CONFIG_SUPPRESS_DUPLICATE_SYMBOL_WARNINGS) :=
 syms-warn-dup-$(CONFIG_ENFORCE_UNIQUE_SYMBOLS) := --error-dup
 
 $(TARGET): TMP = $(@D)/.$(@F).elf32
-$(TARGET): $(TARGET)-syms $(efi-y) boot/mkelf32
+$(TARGET): $(TARGET).mb.efi $(TARGET)-syms $(efi-y) boot/mkelf32
 	./boot/mkelf32 $(notes_phdrs) $(TARGET)-syms $(TMP) $(XEN_IMG_OFFSET) \
 	               `$(NM) $(TARGET)-syms | sed -ne 's/^\([^ ]*\) . __2M_rwdata_end$$/0x\1/p'`
 	od -t x4 -N 8192 $(TMP)  | grep 1badb002 > /dev/null || \
@@ -119,6 +119,11 @@ $(TARGET): $(TARGET)-syms $(efi-y) boot/mkelf32
 		{ echo "No Multiboot2 header found" >&2; false; }
 	mv $(TMP) $(TARGET)
 
+$(TARGET).mb.efi: $(TARGET)-syms
+	$(OBJCOPY) -O binary -S --change-section-address \
+		".efi.pe.header-`$(NM) $(TARGET)-syms | sed -ne 's/^\([^ ]*\) . __image_base__$$/0x\1/p'`" \
+		$(TARGET)-syms $(TARGET).mb.efi
+
 ifneq ($(efi-y),)
 # Check if the compiler supports the MS ABI.
 export XEN_BUILD_EFI := $(shell $(CC) $(XEN_CFLAGS) -c efi/check.c -o efi/check.o 2>/dev/null && echo y)
xen/arch/x86/arch.mk +2 -0
index ce0c1a0e7f..073343d8ea 100644
@@ -7,6 +7,8 @@ CFLAGS += -I$(BASEDIR)/include
 CFLAGS += -I$(BASEDIR)/include/asm-x86/mach-generic
 CFLAGS += -I$(BASEDIR)/include/asm-x86/mach-default
 CFLAGS += -DXEN_IMG_OFFSET=$(XEN_IMG_OFFSET)
+CFLAGS += -DXEN_LOAD_ALIGN=XEN_IMG_OFFSET
+CFLAGS += -DXEN_FILE_ALIGN=0x20
 
 # Prevent floating-point variables from creeping into Xen.
 CFLAGS += -msoft-float
xen/arch/x86/boot/Makefile +1 -0
index 9b31bfcbfb..a559a75e0b 100644
@@ -1,4 +1,5 @@
 obj-bin-y += head.o
+obj-bin-y += pecoff.o
 
 DEFS_H_DEPS = defs.h $(BASEDIR)/include/xen/stdbool.h
 
xen/arch/x86/boot/head.S +1 -0
index 150f7f90a2..2987b4f03a 100644
@@ -1,3 +1,4 @@
+#include <xen/compile.h>
 #include <xen/multiboot.h>
 #include <xen/multiboot2.h>
 #include <public/xen.h>
xen/arch/x86/boot/pecoff.S new file +123 -0

123 lines not on this page · raw message

xen/arch/x86/efi/efi-boot.h +22 -2
index 2541ba1f32..f694a069c9 100644
@@ -32,7 +32,8 @@ static void __init edd_put_string(u8 *dst, size_t n, const char *src)
 }
 #define edd_put_string(d, s) edd_put_string(d, ARRAY_SIZE(d), s)
 
-extern const intpte_t __page_tables_start[], __page_tables_end[];
+extern intpte_t __page_tables_start[], __page_tables_end[];
+
 #define in_page_tables(v) ((intpte_t *)(v) >= __page_tables_start && \
                            (intpte_t *)(v) < __page_tables_end)
 
@@ -568,6 +569,7 @@ static void __init efi_arch_video_init(EFI_GRAPHICS_OUTPUT_PROTOCOL *gop,
 
 static void __init efi_arch_memory_setup(void)
 {
+    intpte_t *pte;
     unsigned int i;
     EFI_STATUS status;
 
@@ -592,6 +594,13 @@ static void __init efi_arch_memory_setup(void)
     if ( !efi_enabled(EFI_LOADER) )
         return;
 
+    if ( efi_enabled(EFI_MB_LOADER) )
+        for ( pte = __page_tables_start; pte < __page_tables_end; pte += ARRAY_SIZE(l2_directmap) )
+            /* Skip relocating the directmap because start_xen() does this for us when
+             * when it updates all superpage-aligned mappings.  */
+            if ( (pte != (intpte_t *)l2_directmap) && (get_pte_flags(*pte) & _PAGE_PRESENT) )
+                *pte += xen_phys_start;
+
     /*
      * Map Xen into the higher mappings, using 2M superpages.
      *
@@ -724,7 +733,18 @@ static bool __init efi_arch_use_config_file(EFI_SYSTEM_TABLE *SystemTable)
 
 static void __init efi_arch_flush_dcache_area(const void *vaddr, UINTN size) { }
 
-void __init efi_multiboot2(EFI_HANDLE ImageHandle, EFI_SYSTEM_TABLE *SystemTable)
+void EFIAPI efi_start(EFI_HANDLE ImageHandle, EFI_SYSTEM_TABLE *SystemTable);
+
+void EFIAPI __init noreturn
+efi_mb_start(EFI_HANDLE ImageHandle, EFI_SYSTEM_TABLE *SystemTable)
+{
+    __set_bit(EFI_MB_LOADER, &efi_flags);
+    efi_start(ImageHandle, SystemTable);
+}
+
+void __init efi_multiboot2(EFI_HANDLE ImageHandle,
+                           EFI_SYSTEM_TABLE *SystemTable,
+                           multiboot2_tag_module_t *dom0_kernel)
 {
     EFI_GRAPHICS_OUTPUT_PROTOCOL *gop;
     UINTN cols, gop_mode = ~0, rows;
xen/arch/x86/efi/stub.c +11 -1
index 9984932626..9bd6355ec3 100644
@@ -15,9 +15,19 @@
  * Here we are in EFI stub. EFI calls are not supported due to lack
  * of relevant functionality in compiler and/or linker.
  *
- * efi_multiboot2() is an exception. Please look below for more details.
+ * efi_mb_start() and efi_multiboot2() are the exceptions.
+ * Please look below for more details.
  */
 
+asm (
+    "    .text                         \n"
+    "    .globl efi_mb_start           \n"
+    "efi_mb_start:                     \n"
+    "    mov    %rcx,%rdi              \n"
+    "    mov    %rdx,%rsi              \n"
+    "    call   efi_multiboot2         \n"
+    );
+
 void __init noreturn efi_multiboot2(EFI_HANDLE ImageHandle,
                                     EFI_SYSTEM_TABLE *SystemTable)
 {
xen/arch/x86/xen.lds.S +34 -0
index 0273f79152..a58ae1e491 100644
@@ -63,7 +63,22 @@ SECTIONS
 
   start_pa = ABSOLUTE(start - __XEN_VIRT_START);
 
+#ifdef EFI
   . = __XEN_VIRT_START + XEN_IMG_OFFSET;
+#else
+  /*
+   * Multiboot2 with an EFI PE/COFF header.
+   *
+   * The PE header must be followed by .text section which
+   * starts at __XEN_VIRT_START + XEN_IMG_OFFSET address.
+   */
+  . = __XEN_VIRT_START + XEN_IMG_OFFSET - efi_pe_head_end + efi_pe_head;
+
+  DECL_SECTION(.efi.pe.header) {
+       *(.efi.pe.header)
+  } :NONE
+#endif
+
   _start = .;
   DECL_SECTION(.text) {
         _stext = .;            /* Text and read-only data */
@@ -289,6 +304,13 @@ SECTIONS
        *(.data.rel)
        *(.data.rel.*)
        CONSTRUCTORS
+       /*
+        * A la the PE/COFF spec, the PE file data section must end at the
+        * alignment boundary equal to FileAlignment in the optional header,
+        * i.e., XEN_FILE_ALIGN.
+        */
+       . = ALIGN(XEN_FILE_ALIGN);
+       __pe_text_raw_end = .;
   } :text
 
   DECL_SECTION(.bss) {
@@ -312,6 +334,8 @@ SECTIONS
   . = ALIGN(SECTION_ALIGN);
   __2M_rwdata_end = .;
 
+  __pe_SizeOfImage = ALIGN(. - __image_base__, MB(16));
+
 #ifdef EFI
   . = ALIGN(4);
   DECL_SECTION(.reloc) {
@@ -341,6 +365,7 @@ SECTIONS
        *(.discard.*)
        *(.eh_frame)
 #ifdef EFI
+       *(.efi.pe.header)
        *(.comment)
        *(.comment.*)
        *(.note.Xen)
@@ -392,5 +417,14 @@ ASSERT((trampoline_end - trampoline_start) < TRAMPOLINE_SPACE - MBI_SPACE_MIN,
 ASSERT((wakeup_stack - wakeup_stack_start) >= WAKEUP_STACK_MIN,
     "wakeup stack too small")
 
+#ifndef EFI
+ASSERT(efi_pe_head_end == _start, "PE header does not end at the beginning of .text section")
+ASSERT(_start == __XEN_VIRT_START + XEN_IMG_OFFSET, ".text section begins at wrong address")
+ASSERT(IS_ALIGNED(_start,      XEN_FILE_ALIGN), "_start misaligned")
+ASSERT(IS_ALIGNED(__bss_start, XEN_FILE_ALIGN), "__bss_start misaligned")
+ASSERT(IS_ALIGNED(__pe_SizeOfImage, XEN_LOAD_ALIGN), "__pe_SizeOfImage is not multiple of XEN_LOAD_ALIGN")
+ASSERT(XEN_LOAD_ALIGN >= XEN_FILE_ALIGN, "XEN_LOAD_ALIGN < XEN_FILE_ALIGN")
+#endif
+
 /* Plenty of boot code assumes that Xen isn't larger than 16M. */
 ASSERT(_end - _start <= MB(16), "Xen too large for early-boot assumptions")
xen/include/xen/efi.h +1 -0
index 94a7e547f9..f7f92b4e3d 100644
@@ -11,6 +11,7 @@ extern unsigned int efi_flags;
 #define EFI_BOOT	0	/* Were we booted from EFI? */
 #define EFI_LOADER	1	/* Were we booted directly from EFI loader? */
 #define EFI_RS		2	/* Can we use runtime services? */
+#define EFI_MB_LOADER	4	/* xen.mb.efi booted directly from EFI loader? */
 
 /* Add fields here only if they need to be referenced from non-EFI code. */
 struct efi {
Signature · 1 line
2.30.0
#6 lore raw

Re: [PATCH v3 2/5] xen/x86: manually build xen.mb.efi binary

Quoted 10 lines · Bobby Eshleman

From: Daniel Kiper <[email protected]>

This patch introduces xen.mb.efi which contains a manually built PE header.

This allows us to support Xen on UEFI Secure Boot-enabled hosts via multiboot2.

xen.mb.efi is a single binary that is loadable by a UEFI loader or via the Multiboot/Multiboot2 protocols.

What's missing here yet very important is why the existing xen.efi doesn't fit and can't be made fit.

Quoted 3 lines

---

Besides (or instead of) the series-wide change log, please have per-patch changes info here.

xen/Makefile endif · line 274

Quoted 12 lines
xen/Makefile +6 -0
@@ -266,29 +266,31 @@ endif
 .PHONY: _build
 _build: $(TARGET)$(CONFIG_XEN_INSTALL_SUFFIX)
 
+define install_xen_links
+	$(INSTALL_DATA) $(TARGET)$1 $(D)$(BOOT_DIR)/$(T)-$(XEN_FULLVERSION)$1
+	ln -f -s $(T)-$(XEN_FULLVERSION)$1 $(D)$(BOOT_DIR)/$(T)-$(XEN_VERSION).$(XEN_SUBVERSION)$1
+	ln -f -s $(T)-$(XEN_FULLVERSION)$1 $(D)$(BOOT_DIR)/$(T)-$(XEN_VERSION)$1
+	ln -f -s $(T)-$(XEN_FULLVERSION)$1 $(D)$(BOOT_DIR)/$(T)$1
+endef

If you abstract this away, please take the opportunity to fold "-f -s" into a single option.

xen/Makefile endif · line 283

Quoted 12 lines
xen/Makefile +2 -4
 .PHONY: _install
 _install: D=$(DESTDIR)
 _install: T=$(notdir $(TARGET))
 _install: Z=$(CONFIG_XEN_INSTALL_SUFFIX)
 _install: $(TARGET)$(CONFIG_XEN_INSTALL_SUFFIX)
 	[ -d $(D)$(BOOT_DIR) ] || $(INSTALL_DIR) $(D)$(BOOT_DIR)
-	$(INSTALL_DATA) $(TARGET)$(Z) $(D)$(BOOT_DIR)/$(T)-$(XEN_FULLVERSION)$(Z)
-	ln -f -s $(T)-$(XEN_FULLVERSION)$(Z) $(D)$(BOOT_DIR)/$(T)-$(XEN_VERSION).$(XEN_SUBVERSION)$(Z)
-	ln -f -s $(T)-$(XEN_FULLVERSION)$(Z) $(D)$(BOOT_DIR)/$(T)-$(XEN_VERSION)$(Z)
-	ln -f -s $(T)-$(XEN_FULLVERSION)$(Z) $(D)$(BOOT_DIR)/$(T)$(Z)
+	$(call install_xen_links,$(Z))
+	$(call install_xen_links,.mb.efi)

This is common code, so will affect Arm as well. I don't think your addition can be unconditional.

Quoted 7 lines
xen/Makefile +0 -1
 	[ -d "$(D)$(DEBUG_DIR)" ] || $(INSTALL_DIR) $(D)$(DEBUG_DIR)
 	$(INSTALL_DATA) $(TARGET)-syms $(D)$(DEBUG_DIR)/$(T)-syms-$(XEN_FULLVERSION)
 	$(INSTALL_DATA) $(TARGET)-syms.map $(D)$(DEBUG_DIR)/$(T)-syms-$(XEN_FULLVERSION).map
 	$(INSTALL_DATA) $(KCONFIG_CONFIG) $(D)$(BOOT_DIR)/$(T)-$(XEN_FULLVERSION).config
 	if [ -r $(TARGET).efi -a -n '$(EFI_DIR)' ]; then \
 		[ -d $(D)$(EFI_DIR) ] || $(INSTALL_DIR) $(D)$(EFI_DIR); \
-		$(INSTALL_DATA) $(TARGET).efi $(D)$(EFI_DIR)/$(T)-$(XEN_FULLVERSION).efi; \

xen/Makefile endif · line 296

Quoted 10 lines
xen/Makefile +1 -3
 		if [ -e $(TARGET).efi.map ]; then \
 			$(INSTALL_DATA) $(TARGET).efi.map $(D)$(DEBUG_DIR)/$(T)-$(XEN_FULLVERSION).efi.map; \
 		fi; \
-		ln -sf $(T)-$(XEN_FULLVERSION).efi $(D)$(EFI_DIR)/$(T)-$(XEN_VERSION).$(XEN_SUBVERSION).efi; \
-		ln -sf $(T)-$(XEN_FULLVERSION).efi $(D)$(EFI_DIR)/$(T)-$(XEN_VERSION).efi; \
-		ln -sf $(T)-$(XEN_FULLVERSION).efi $(D)$(EFI_DIR)/$(T).efi; \
+		$(call install_xen_links,.efi)) \
 		if [ -n '$(EFI_MOUNTPOINT)' -a -n '$(EFI_VENDOR)' ]; then \
 			$(INSTALL_DATA) $(TARGET).efi $(D)$(EFI_MOUNTPOINT)/efi/$(EFI_VENDOR)/$(T)-$(XEN_FULLVERSION).efi; \
 		elif [ "$(D)" = "$(patsubst $(shell cd $(XEN_ROOT) && pwd)/%,%,$(D))" ]; then \

Since this part of the patch is a non-negligible fraction of the patch and since this installation step doesn't need to be an integral part of the change, may I suggest / ask that you split this off into a separate change? Possibly the installing of the new binary could remain here, but then the breaking out of the install_xen_links macro (which imo also would better use dashes in place of the underscores) could still be factored out.

xen/arch/x86/Makefile syms-warn-dup-$(CONFIG_SUPPRESS_DUPLICATE_SYMBOL_WARNINGS) := · line 113

Quoted 8 lines
xen/arch/x86/Makefile +1 -1
@@ -110,7 +110,7 @@ syms-warn-dup-$(CONFIG_SUPPRESS_DUPLICATE_SYMBOL_WARNINGS) :=
 syms-warn-dup-$(CONFIG_ENFORCE_UNIQUE_SYMBOLS) := --error-dup
 
 $(TARGET): TMP = $(@D)/.$(@F).elf32
-$(TARGET): $(TARGET)-syms $(efi-y) boot/mkelf32
+$(TARGET): $(TARGET).mb.efi $(TARGET)-syms $(efi-y) boot/mkelf32

While perhaps mostly cosmetic, I'd prefer additions to be done after the existing (pseudo-)dependencies, not as the very first item. $(TARGET)-syms still is the main dependency here, and it should remain this way.

Speaking of (pseudo-)dependencies - I was hoping that we could avoid further extending this sub-optimal approach.

xen/arch/x86/Makefile $(TARGET): $(TARGET)-syms $(efi-y) boot/mkelf32 · line 125

Quoted 8 lines
xen/arch/x86/Makefile +4 -0
@@ -119,6 +119,11 @@ $(TARGET): $(TARGET)-syms $(efi-y) boot/mkelf32
 		{ echo "No Multiboot2 header found" >&2; false; }
 	mv $(TMP) $(TARGET)
 
+$(TARGET).mb.efi: $(TARGET)-syms
+	$(OBJCOPY) -O binary -S --change-section-address \
+		".efi.pe.header-`$(NM) $(TARGET)-syms | sed -ne 's/^\([^ ]*\) . __image_base__$$/0x\1/p'`" \
+		$(TARGET)-syms $(TARGET).mb.efi

The quoting is very hard to follow here. While using the shell's $() would already seem to be an improvement, I don't see why you shouldn't be able to have make construct the tail of the section name by using $(shell ...). This way, in case of someone needing to debug this, the resulting command line would be more explict.

I have to admit I could also do with a few words in the description as to what this playing with a specific section's address is actually needed for, and how it's guaranteed that this isn't going to end in confusion (e.g. because of trying to put the section at where other stuff is already sitting, perhaps just partially). It's also unclear to me why the new address is calculated by subtracting the image base address. The PE file header is, aiui, assumed to live at RVA 0, i.e. precisely at the image base.

Further - why the -S? xen.efi comes with a proper symbol table.

And finally I'm not convinced of it being a good idea to use __image_base__ here - that symbol exists only to help the linker script cover both ELF and PE binaries. It would be good is new road blocks towards eliminating this crutch could be avoided. Can't you e.g. get the main program header's specified address and subtract XEN_IMG_OFFSET?

xen/arch/x86/arch.mk CFLAGS += -I$(BASEDIR)/include · line 11

Quoted 8 lines
xen/arch/x86/arch.mk +2 -0
@@ -7,6 +7,8 @@ CFLAGS += -I$(BASEDIR)/include
 CFLAGS += -I$(BASEDIR)/include/asm-x86/mach-generic
 CFLAGS += -I$(BASEDIR)/include/asm-x86/mach-default
 CFLAGS += -DXEN_IMG_OFFSET=$(XEN_IMG_OFFSET)
+CFLAGS += -DXEN_LOAD_ALIGN=XEN_IMG_OFFSET
+CFLAGS += -DXEN_FILE_ALIGN=0x20

The former is merely coincidence - I don't think you want to use an offset value as alignment. For both of them I think once you go this far, you also want to consolidate with xen.efi's --section-alignment= and --file-alignment= settings, such that the values don't need to be kept in sync "manually".

What I could see is deriving XEN_IMG_OFFSET from a hypothetical XEN_SECTION_ALIGN value, because we indeed want the first section (.text) to start at the 2nd large page from the image base.

xen/arch/x86/boot/head.S line 4

Quoted 7 lines
xen/arch/x86/boot/head.S +1 -0
@@ -1,3 +1,4 @@
+#include <xen/compile.h>
 #include <xen/multiboot.h>
 #include <xen/multiboot2.h>
 #include <public/xen.h>

Why?

xen/arch/x86/boot/pecoff.S line 8

Quoted 11 lines
xen/arch/x86/boot/pecoff.S new file +8 -0
@@ -0,0 +1,123 @@
+#include <xen/compile.h>
+#include <asm/page.h>
+
+#define sym_offs(sym)     ((sym) - __XEN_VIRT_START)
+
+        .section .efi.pe.header, "a", @progbits
+
+GLOBAL(efi_pe_head)

I don't think this should be global. But I'll also comment on the linker script part using it. In any event there you only care about efi_pe_head - efi_pe_head_end, i.e. the size of this section. Linker scripts have SIZEOF() for this purpose - is it not possible to use that here?

xen/arch/x86/boot/pecoff.S line 17

Quoted 9 lines
xen/arch/x86/boot/pecoff.S new file +9 -0
+        /*
+         * Legacy EXE header.
+         *
+         * Most of it is copied from binutils package, version 2.30,
+         * include/coff/pe.h:struct external_PEI_filehdr and
+         * bfd/peXXigen.c:_bfd_XXi_only_swap_filehdr_out().
+         *
+         * Page is equal 512 bytes here.
+         * Paragraph is equal 16 bytes here.

"is equal" is not very clear imo. How about '"Page" refers to an aligned block of 512 bytes here'?

Quoted 12 lines
xen/arch/x86/boot/pecoff.S new file +12 -0
+         */
+        .short  0x5a4d                               /* EXE magic number. */
+        .short  0x90                                 /* Bytes on last page of file. */
+        .short  0x3                                  /* Pages in file. */
+        .short  0                                    /* Relocations. */
+        .short  0x4                                  /* Size of header in paragraphs. */
+        .short  0                                    /* Minimum extra paragraphs needed. */
+        .short  0xffff                               /* Maximum extra paragraphs needed. */
+        .short  0                                    /* Initial (relative) SS value. */
+        .short  0xb8                                 /* Initial SP value. */
+        .short  0                                    /* Checksum. */
+        .short  0                                    /* Initial IP value. */

xen/arch/x86/boot/pecoff.S line 39

Quoted 10 lines
xen/arch/x86/boot/pecoff.S new file +10 -0
+        .short  0                                    /* Initial (relative) CS value. */
+        .short  0x40                                 /* File address of relocation table. */
+        .short  0                                    /* Overlay number. */
+        .fill   4, 2, 0                              /* Reserved words. */
+        .short  0                                    /* OEM identifier. */
+        .short  0                                    /* OEM information. */
+        .fill   10, 2, 0                             /* Reserved words. */
+        .long   Lpe_header - efi_pe_head             /* File address of the PE header. */
+
+Lpe_header:

Was this meant to have a leading '.' (also again further down)? Else I don't see what the uppercase L is about.

xen/arch/x86/boot/pecoff.S line 50

Quoted 11 lines
xen/arch/x86/boot/pecoff.S new file +11 -0
+        /*
+         * PE/COFF header.
+         *
+         * The PE/COFF format is defined by Microsoft, and is available from
+         * https://docs.microsoft.com/en-us/windows/win32/debug/pe-format
+         *
+         * Some ideas are taken from Linux kernel and Xen ARM64.
+         */
+        .ascii  "PE\0\0"                             /* PE signature. */
+        .short  0x8664                               /* Machine: IMAGE_FILE_MACHINE_AMD64 */
+        .short  1                                    /* NumberOfSections. */

So like in xen-syms / xen.gz everything gets munged into a single section? Not very nice, I would say.

xen/arch/x86/boot/pecoff.S line 51

Quoted 1 line
xen/arch/x86/boot/pecoff.S new file +1 -0
+        .long   XEN_COMPILE_POSIX_TIME               /* TimeDateStamp. */

This wants to honor SOURCE_DATE_EPOCH (where for xen.efi we pass --no-insert-timestamp to the linker). Perhaps a missed re-base?

xen/arch/x86/boot/pecoff.S line 54

Quoted 3 lines
xen/arch/x86/boot/pecoff.S new file +3 -0
+        .long   0                                    /* PointerToSymbolTable. */
+        .long   0                                    /* NumberOfSymbols. */
+        .short  Lsection_table - Loptional_header      /* SizeOfOptionalHeader. */

Nit: Too many blanks before the comment.

xen/arch/x86/boot/pecoff.S line 60

Quoted 6 lines
xen/arch/x86/boot/pecoff.S new file +6 -0
+        .short  0x226                                /* Characteristics:
+                                                      *   IMAGE_FILE_EXECUTABLE_IMAGE |
+                                                      *   IMAGE_FILE_LARGE_ADDRESS_AWARE |
+                                                      *   IMAGE_FILE_DEBUG_STRIPPED |
+                                                      *   IMAGE_FILE_LINE_NUMS_STRIPPED
+                                                      */

You don't specify IMAGE_FILE_RELOCS_STRIPPED here, but you also don't seem to generate base relocations. How is this going to work?

xen/arch/x86/boot/pecoff.S line 68

Quoted 7 lines
xen/arch/x86/boot/pecoff.S new file +7 -0
+Loptional_header:
+        .short  0x20b                                /* PE format: PE32+ */
+        .byte   0                                    /* MajorLinkerVersion. */
+        .byte   0                                    /* MinorLinkerVersion. */
+        .long   __2M_rwdata_end - efi_pe_head_end    /* SizeOfCode. */
+        .long   0                                    /* SizeOfInitializedData. */
+        .long   0                                    /* SizeOfUninitializedData. */

Everything's code?

xen/arch/x86/boot/pecoff.S line 71

Quoted 3 lines
xen/arch/x86/boot/pecoff.S new file +3 -0
+        .long   sym_offs(efi_mb_start)               /* AddressOfEntryPoint. */
+        .long   sym_offs(start)                      /* BaseOfCode. */
+        .quad   sym_offs(__image_base__)             /* ImageBase. */

This last value is zero, isn't it? Can a PE image validly live at address 0? I have to admit that I question all of the sym_offs() uses here and below, which goes along with the lack of base relocations mentioned above.

xen/arch/x86/boot/pecoff.S line 81

Quoted 10 lines
xen/arch/x86/boot/pecoff.S new file +10 -0
+        .long   XEN_LOAD_ALIGN                       /* SectionAlignment. */
+        .long   XEN_FILE_ALIGN                       /* FileAlignment. */
+        .short  2                                    /* MajorOperatingSystemVersion. */
+        .short  0                                    /* MinorOperatingSystemVersion. */
+        .short  XEN_VERSION                          /* MajorImageVersion. */
+        .short  XEN_SUBVERSION                       /* MinorImageVersion. */
+        .short  2                                    /* MajorSubsystemVersion. */
+        .short  0                                    /* MinorSubsystemVersion. */
+        .long   0                                    /* Win32VersionValue. */
+        .long   __pe_SizeOfImage                     /* SizeOfImage. */

I'm not convinced of the utility of how you calculate this value just to use it here. Right now the value has to be MB(16) - any smaller value will cause breakage.

Quoted 8 lines
xen/arch/x86/boot/pecoff.S new file +8 -0
+        .long   efi_pe_head_end - efi_pe_head        /* SizeOfHeaders. */
+        .long   0                                    /* CheckSum. */
+        .short  0xa                                  /* Subsystem: EFI application. */
+        .short  0                                    /* DllCharacteristics. */
+        .quad   0                                    /* SizeOfStackReserve. */
+        .quad   0                                    /* SizeOfStackCommit. */
+        .quad   0                                    /* SizeOfHeapReserve. */
+        .quad   0                                    /* SizeOfHeapCommit. */

xen/arch/x86/boot/pecoff.S line 99

Quoted 10 lines
xen/arch/x86/boot/pecoff.S new file +10 -0
+        .long   0                                    /* LoaderFlags. */
+        .long   0x6                                  /* NumberOfRvaAndSizes. */
+
+        /* Data Directories. */
+        .quad   0                                    /* Export Table. */
+        .quad   0                                    /* Import Table. */
+        .quad   0                                    /* Resource Table. */
+        .quad   0                                    /* Exception Table. */
+        .quad   0                                    /* Certificate Table. */
+        .quad   0                                    /* Base Relocation Table. */

Based on what was the number of directory entries chosen here? 6 is a pretty unusual value - typically it would be 16, I think. I'm fine if this is for space savings (and known to be compatible), but then why not strip the other unused ones as well? Then again for the build ID don't you need the 7th entry (see xen.efi)? Or are you intentionally not exposing it in the PE way (in which case saying so, and why, would be needed in the description)?

xen/arch/x86/boot/pecoff.S line 106

Quoted 6 lines
xen/arch/x86/boot/pecoff.S new file +6 -0
+Lsection_table:
+        .ascii  ".text\0\0\0"                        /* Name. */
+        .long   __2M_rwdata_end - efi_pe_head_end    /* VirtualSize. */
+        .long   sym_offs(start)                      /* VirtualAddress. */
+        .long   __pe_text_raw_end - efi_pe_head_end  /* SizeOfRawData. */
+        .long   efi_pe_head_end - efi_pe_head        /* PointerToRawData. */

Isn't this a file offset? If so, can it legitimately and reliably be calculated by a difference of two addresses?

xen/arch/x86/boot/pecoff.S line 117

Quoted 11 lines
xen/arch/x86/boot/pecoff.S new file +11 -0
+        .long   0                                    /* PointerToRelocations. */
+        .long   0                                    /* PointerToLinenumbers. */
+        .short  0                                    /* NumberOfRelocations. */
+        .short  0                                    /* NumberOfLinenumbers. */
+        .long   0xe0500020                           /* Characteristics:
+                                                      *   IMAGE_SCN_CNT_CODE |
+                                                      *   IMAGE_SCN_ALIGN_16BYTES |
+                                                      *   IMAGE_SCN_MEM_EXECUTE |
+                                                      *   IMAGE_SCN_MEM_READ |
+                                                      *   IMAGE_SCN_MEM_WRITE
+                                                      */

At least the alignment specification here is fake. I realize it doesn't matter for loading purposes, but if an arbirary value was chosen it should imo be said so in a comment, to avoid future readers wondering.

xen/arch/x86/efi/efi-boot.h static void __init edd_put_string(u8 *dst, size_t n, const char *src) · line 35

Quoted 8 lines
xen/arch/x86/efi/efi-boot.h +1 -1
@@ -32,7 +32,8 @@ static void __init edd_put_string(u8 *dst, size_t n, const char *src)
 }
 #define edd_put_string(d, s) edd_put_string(d, ARRAY_SIZE(d), s)
 
-extern const intpte_t __page_tables_start[], __page_tables_end[];
+extern intpte_t __page_tables_start[], __page_tables_end[];

I'm afraid I'm against this, no matter that it may be difficult to do differently what you do below. IOW ...

Quoted 9 lines
xen/arch/x86/efi/efi-boot.h +1 -0
@@ -568,6 +569,7 @@ static void __init efi_arch_video_init(EFI_GRAPHICS_OUTPUT_PROTOCOL *gop,
 
 static void __init efi_arch_memory_setup(void)
 {
+    intpte_t *pte;
     unsigned int i;
     EFI_STATUS status;
 

xen/arch/x86/efi/efi-boot.h static void __init efi_arch_memory_setup(void) · line 602

Quoted 9 lines
xen/arch/x86/efi/efi-boot.h +6 -0
@@ -592,6 +594,13 @@ static void __init efi_arch_memory_setup(void)
     if ( !efi_enabled(EFI_LOADER) )
         return;
 
+    if ( efi_enabled(EFI_MB_LOADER) )
+        for ( pte = __page_tables_start; pte < __page_tables_end; pte += ARRAY_SIZE(l2_directmap) )
+            /* Skip relocating the directmap because start_xen() does this for us when
+             * when it updates all superpage-aligned mappings.  */
+            if ( (pte != (intpte_t *)l2_directmap) && (get_pte_flags(*pte) & _PAGE_PRESENT) )
+                *pte += xen_phys_start;

... I consider this an RFC hack for which a clean solution wants to be found (note how __setup_arch() gets away without such). Also nit: Comment style.

Quoted 8 lines
xen/arch/x86/efi/efi-boot.h +3 -1
@@ -724,7 +733,18 @@ static bool __init efi_arch_use_config_file(EFI_SYSTEM_TABLE *SystemTable)
 
 static void __init efi_arch_flush_dcache_area(const void *vaddr, UINTN size) { }
 
-void __init efi_multiboot2(EFI_HANDLE ImageHandle, EFI_SYSTEM_TABLE *SystemTable)
+void EFIAPI efi_start(EFI_HANDLE ImageHandle, EFI_SYSTEM_TABLE *SystemTable);
+
+void EFIAPI __init noreturn

xen/arch/x86/efi/efi-boot.h static bool __init efi_arch_use_config_file(EFI_SYSTEM_TABLE *SystemTable) · line 748

Quoted 10 lines
xen/arch/x86/efi/efi-boot.h +9 -0
+efi_mb_start(EFI_HANDLE ImageHandle, EFI_SYSTEM_TABLE *SystemTable)
+{
+    __set_bit(EFI_MB_LOADER, &efi_flags);
+    efi_start(ImageHandle, SystemTable);
+}
+
+void __init efi_multiboot2(EFI_HANDLE ImageHandle,
+                           EFI_SYSTEM_TABLE *SystemTable,
+                           multiboot2_tag_module_t *dom0_kernel)
 {

Hmm, yet another entry point. See also at the very bottom.

Quoted 9 lines
xen/arch/x86/efi/stub.c +2 -1
@@ -15,9 +15,19 @@
  * Here we are in EFI stub. EFI calls are not supported due to lack
  * of relevant functionality in compiler and/or linker.
  *
- * efi_multiboot2() is an exception. Please look below for more details.
+ * efi_mb_start() and efi_multiboot2() are the exceptions.
+ * Please look below for more details.

xen/arch/x86/efi/stub.c line 29

Quoted 10 lines
xen/arch/x86/efi/stub.c +8 -0
  */
 
+asm (
+    "    .text                         \n"
+    "    .globl efi_mb_start           \n"
+    "efi_mb_start:                     \n"
+    "    mov    %rcx,%rdi              \n"
+    "    mov    %rdx,%rsi              \n"
+    "    call   efi_multiboot2         \n"
+    );

Okay, this I understand is for calling conventions translation. A comment saying so would be nice. Plus I don't see why this then uses "call", not "jmp".

Quoted 11 lines
xen/arch/x86/xen.lds.S +4 -0
@@ -63,7 +63,22 @@ SECTIONS
 
   start_pa = ABSOLUTE(start - __XEN_VIRT_START);
 
+#ifdef EFI
   . = __XEN_VIRT_START + XEN_IMG_OFFSET;
+#else
+  /*
+   * Multiboot2 with an EFI PE/COFF header.

xen/arch/x86/xen.lds.S SECTIONS · line 80

Quoted 10 lines
xen/arch/x86/xen.lds.S +10 -0
+   *
+   * The PE header must be followed by .text section which
+   * starts at __XEN_VIRT_START + XEN_IMG_OFFSET address.
+   */
+  . = __XEN_VIRT_START + XEN_IMG_OFFSET - efi_pe_head_end + efi_pe_head;
+
+  DECL_SECTION(.efi.pe.header) {
+       *(.efi.pe.header)
+  } :NONE
+#endif

"Must be followed" in the comment is about file layout, not address layout. Yet the latter is what matters in the linker script. If there is a true requirement for it to be exactly like this, more explanation is needed in the comment. But it seems more likely to me that this simply isn't correct. As said elsewhere, the executable header of a PE image lives at RVA 0 afaict.

xen/arch/x86/xen.lds.S SECTIONS · line 314

Quoted 12 lines
xen/arch/x86/xen.lds.S +7 -0
@@ -289,6 +304,13 @@ SECTIONS
        *(.data.rel)
        *(.data.rel.*)
        CONSTRUCTORS
+       /*
+        * A la the PE/COFF spec, the PE file data section must end at the
+        * alignment boundary equal to FileAlignment in the optional header,
+        * i.e., XEN_FILE_ALIGN.
+        */
+       . = ALIGN(XEN_FILE_ALIGN);
+       __pe_text_raw_end = .;
   } :text

What is a "PE file data section"? Yes, the file size of a section must be a multiple of the specified file alignment. With the present value of 32 bytes this isn't much of an issue, but already in case we were in need of going up to 512 bytes I'd say this is undue overhead for the ELF image.

I could see you not advancing . here (by using

       __pe_text_raw_end = ALIGN(XEN_FILE_ALIGN);

) and then making sure the generated image gets padded as necessary.

xen/arch/x86/xen.lds.S ASSERT((trampoline_end - trampoline_start) < TRAMPOLINE_SPACE - MBI_SPACE_MIN, · line 421

Quoted 6 lines
xen/arch/x86/xen.lds.S +2 -0
@@ -392,5 +417,14 @@ ASSERT((trampoline_end - trampoline_start) < TRAMPOLINE_SPACE - MBI_SPACE_MIN,
 ASSERT((wakeup_stack - wakeup_stack_start) >= WAKEUP_STACK_MIN,
     "wakeup stack too small")
 
+#ifndef EFI
+ASSERT(efi_pe_head_end == _start, "PE header does not end at the beginning of .text section")

As said earlier - I question this relationship.

xen/arch/x86/xen.lds.S ASSERT((trampoline_end - trampoline_start) < TRAMPOLINE_SPACE - MBI_SPACE_MIN, · line 422

Quoted 1 line
xen/arch/x86/xen.lds.S +1 -0
+ASSERT(_start == __XEN_VIRT_START + XEN_IMG_OFFSET, ".text section begins at wrong address")

I'd then hope this could go away as well.

xen/arch/x86/xen.lds.S ASSERT((trampoline_end - trampoline_start) < TRAMPOLINE_SPACE - MBI_SPACE_MIN, · line 423

Quoted 1 line
xen/arch/x86/xen.lds.S +1 -0
+ASSERT(IS_ALIGNED(_start,      XEN_FILE_ALIGN), "_start misaligned")

I can't see how this could trigger when the former one doesn't.

xen/arch/x86/xen.lds.S ASSERT((trampoline_end - trampoline_start) < TRAMPOLINE_SPACE - MBI_SPACE_MIN, · line 424

Quoted 1 line
xen/arch/x86/xen.lds.S +1 -0
+ASSERT(IS_ALIGNED(__bss_start, XEN_FILE_ALIGN), "__bss_start misaligned")

What is this trying to verify?

xen/arch/x86/xen.lds.S ASSERT((trampoline_end - trampoline_start) < TRAMPOLINE_SPACE - MBI_SPACE_MIN, · line 425

Quoted 1 line
xen/arch/x86/xen.lds.S +1 -0
+ASSERT(IS_ALIGNED(__pe_SizeOfImage, XEN_LOAD_ALIGN), "__pe_SizeOfImage is not multiple of XEN_LOAD_ALIGN")

This looks odd too, but I've commented on __pe_SizeOfImage further up already anyway.

xen/arch/x86/xen.lds.S ASSERT((trampoline_end - trampoline_start) < TRAMPOLINE_SPACE - MBI_SPACE_MIN, · line 426

Quoted 1 line
xen/arch/x86/xen.lds.S +1 -0
+ASSERT(XEN_LOAD_ALIGN >= XEN_FILE_ALIGN, "XEN_LOAD_ALIGN < XEN_FILE_ALIGN")

Why? I would generally consider the two values pretty much independent.

xen/include/xen/efi.h extern unsigned int efi_flags; · line 14

Quoted 7 lines
xen/include/xen/efi.h +1 -0
@@ -11,6 +11,7 @@ extern unsigned int efi_flags;
 #define EFI_BOOT	0	/* Were we booted from EFI? */
 #define EFI_LOADER	1	/* Were we booted directly from EFI loader? */
 #define EFI_RS		2	/* Can we use runtime services? */
+#define EFI_MB_LOADER	4	/* xen.mb.efi booted directly from EFI loader? */

Is a separate flag really needed? I realize this is connected to the page table relocation approach, so might go away anyway.

Jan

#7 lore raw

Re: [PATCH v3 2/5] xen/x86: manually build xen.mb.efi binary

Jan,

I mulled over your feedback and I think I can now see your reservations with this series.

I'm wondering if the long-term goal of using the xen mb1/mb2 binary as the basis for creating a EFI loadable mb1/mb2 payload is actually the wrong approach.

After all, I do not see a feasible way to maintain the comprehensive sectioning, the proper reloc table, the proper debug directory, etc... that is found in the current xen.efi using the approach in this series, which would mean maintaining a third binary forever.

What is your intuition WRT the idea that instead of trying add a PE/COFF hdr in front of Xen's mb2 bin, we instead go the route of introducing valid mb2 entry points into xen.efi?

At the end of the day, our goal is just to have a binary that meets these requirements:

* Is verifiable with shim (PE/COFF)
* May boot on BIOS platforms via grub2
* May boot on EFI platforms via grub2 or EFI loader

Thanks

Signature · 2 lines
Bobby Eshleman
SE at Vates SAS
#8 lore raw

Re: [PATCH v3 2/5] xen/x86: manually build xen.mb.efi binary

Quoted 3 lines · Bob Eshleman

What is your intuition WRT the idea that instead of trying add a PE/COFF hdr in front of Xen's mb2 bin, we instead go the route of introducing valid mb2 entry points into xen.efi?

At the first glance I think this is going to be less intrusive, and hence to be preferred. But of course I haven't experimented in any way ...

Jan

#9 lore raw

Re: [PATCH v3 2/5] xen/x86: manually build xen.mb.efi binary

Quoted 7 lines · Jan Beulich
Quoted 3 lines · Bob Eshleman

What is your intuition WRT the idea that instead of trying add a PE/COFF hdr in front of Xen's mb2 bin, we instead go the route of introducing valid mb2 entry points into xen.efi?

At the first glance I think this is going to be less intrusive, and hence to be preferred. But of course I haven't experimented in any way ...

When I worked on this a few years ago I tried that way. Sadly I failed because I was not able to produce "linear" PE image using binutils exiting that days. Though I think you should try once again. Maybe newer binutils are more flexible and will be able to produce a PE image with properties required by Multiboot2 protocol.

Daniel

#10 lore raw

Re: [PATCH v3 2/5] xen/x86: manually build xen.mb.efi binary

Quoted 6 lines · Daniel Kiper
Quoted 5 lines · Jan Beulich
Quoted 3 lines · Bob Eshleman

What is your intuition WRT the idea that instead of trying add a PE/COFF hdr in front of Xen's mb2 bin, we instead go the route of introducing valid mb2 entry points into xen.efi?

Quoted 6 lines
Quoted 2 lines

At the first glance I think this is going to be less intrusive, and hence to be preferred. But of course I haven't experimented in any way ...

When I worked on this a few years ago I tried that way. Sadly I failed because I was not able to produce "linear" PE image using binutils exiting that days.

What is a "linear" PE image?

Quoted 3 lines

Maybe
newer binutils are more flexible and will be able to produce a PE image with properties required by Multiboot2 protocol.

Isn't all you need the MB2 header within the first so many bytes of the (disk) image? Or was it the image as loaded into memory? Both should be possible to arrange for.

Jan

#11 lore raw

Re: [PATCH v3 2/5] xen/x86: manually build xen.mb.efi binary

Quoted 7 lines · Jan Beulich
Quoted 6 lines · Daniel Kiper
Quoted 5 lines · Jan Beulich
Quoted 3 lines · Bob Eshleman

What is your intuition WRT the idea that instead of trying add a PE/COFF hdr in front of Xen's mb2 bin, we instead go the route of introducing valid mb2 entry points into xen.efi?

Quoted 8 lines
Quoted 6 lines
Quoted 2 lines

At the first glance I think this is going to be less intrusive, and hence to be preferred. But of course I haven't experimented in any way ...

When I worked on this a few years ago I tried that way. Sadly I failed because I was not able to produce "linear" PE image using binutils exiting that days.

What is a "linear" PE image?

The problem with Multiboot family protocols is that all code and data sections have to be glued together in the image and as such loaded into the memory (IIRC BSS is an exception but it has to live behind the image). So, you cannot use PE image which has different representation in file and memory. IIRC by default at least code and data sections in xen.efi have different sizes in PE file and in memory. I tried to fix that using linker script and objcopy but it did not work. Sadly I do not remember the details but there is pretty good chance you can find relevant emails in Xen-devel archive with me explaining what kind of problems I met.

Quoted 7 lines
Quoted 3 lines

Maybe
newer binutils are more flexible and will be able to produce a PE image with properties required by Multiboot2 protocol.

Isn't all you need the MB2 header within the first so many bytes of the (disk) image? Or was it the image as loaded into memory? Both should be possible to arrange for.

IIRC Multiboot2 protocol requires the header in first 32 kiB of an image. So, this is not a problem.

Daniel

#12 lore raw

Re: [PATCH v3 2/5] xen/x86: manually build xen.mb.efi binary

Quoted 17 lines · Daniel Kiper
Quoted 7 lines · Jan Beulich
Quoted 6 lines · Daniel Kiper
Quoted 5 lines · Jan Beulich
Quoted 3 lines · Bob Eshleman

What is your intuition WRT the idea that instead of trying add a PE/COFF hdr in front of Xen's mb2 bin, we instead go the route of introducing valid mb2 entry points into xen.efi?

Quoted 8 lines
Quoted 6 lines
Quoted 2 lines

At the first glance I think this is going to be less intrusive, and hence to be preferred. But of course I haven't experimented in any way ...

When I worked on this a few years ago I tried that way. Sadly I failed because I was not able to produce "linear" PE image using binutils exiting that days.

What is a "linear" PE image?

Quoted 10 lines

The problem with Multiboot family protocols is that all code and data sections have to be glued together in the image and as such loaded into the memory (IIRC BSS is an exception but it has to live behind the image). So, you cannot use PE image which has different representation in file and memory. IIRC by default at least code and data sections in xen.efi have different sizes in PE file and in memory. I tried to fix that using linker script and objcopy but it did not work. Sadly I do not remember the details but there is pretty good chance you can find relevant emails in Xen-devel archive with me explaining what kind of problems I met.

Ah, this rings a bell. Even the .bss-is-last assumption doesn't hold, because .reloc (for us as well as in general) comes later, but needs loading (in the right place). Since even xen.gz isn't simply the compressed linker output, but a post-processed (by mkelf32) image, maybe what we need is a build tool doing similar post-processing on xen.efi? Otoh getting disk image and in-memory image aligned ought to be possible by setting --section-alignment= and --file-alignment= to the same value (resulting in a much larger file) - adjusting file positions would effectively be what a post-processing tool would need to do (like with mkelf32 perhaps we could then at least save the first ~2Mb of space). Which would still leave .reloc to be dealt with
- maybe we could place this after .init, but still ahead of __init_end (such that the memory would get freed late in the boot process). Not sure whether EFI loaders would "like" such an unusual placement.

Also not sure what to do with Dwarf debug info, which just recently we managed to avoid needing to strip unconditionally.

Quoted 10 lines
Quoted 7 lines
Quoted 3 lines

Maybe
newer binutils are more flexible and will be able to produce a PE image with properties required by Multiboot2 protocol.

Isn't all you need the MB2 header within the first so many bytes of the (disk) image? Or was it the image as loaded into memory? Both should be possible to arrange for.

IIRC Multiboot2 protocol requires the header in first 32 kiB of an image. So, this is not a problem.

I was about to ask "Disk image or in-memory image?" But this won't matter if the image as a whole got linearized, as long as the first section doesn't start to high up. I notice that xen-syms doesn't fit this requirement either, only the output of mkelf32 does. Which suggests that there may not be a way around a post-processing tool.

Jan

#13 lore raw

Re: [PATCH v3 2/5] xen/x86: manually build xen.mb.efi binary

Quoted 29 lines · Jan Beulich
Quoted 17 lines · Daniel Kiper
Quoted 7 lines · Jan Beulich
Quoted 6 lines · Daniel Kiper
Quoted 5 lines · Jan Beulich
Quoted 3 lines · Bob Eshleman

What is your intuition WRT the idea that instead of trying add a PE/COFF hdr in front of Xen's mb2 bin, we instead go the route of introducing valid mb2 entry points into xen.efi?

Quoted 8 lines
Quoted 6 lines
Quoted 2 lines

At the first glance I think this is going to be less intrusive, and hence to be preferred. But of course I haven't experimented in any way ...

When I worked on this a few years ago I tried that way. Sadly I failed because I was not able to produce "linear" PE image using binutils exiting that days.

What is a "linear" PE image?

Quoted 10 lines

The problem with Multiboot family protocols is that all code and data sections have to be glued together in the image and as such loaded into the memory (IIRC BSS is an exception but it has to live behind the image). So, you cannot use PE image which has different representation in file and memory. IIRC by default at least code and data sections in xen.efi have different sizes in PE file and in memory. I tried to fix that using linker script and objcopy but it did not work. Sadly I do not remember the details but there is pretty good chance you can find relevant emails in Xen-devel archive with me explaining what kind of problems I met.

Quoted 3 lines

Ah, this rings a bell. Even the .bss-is-last assumption doesn't hold, because .reloc (for us as well as in general) comes later, but needs loading (in the right place). Since even xen.gz isn't simply the

However, IIRC it is not used when Xen is loaded through Multiboot2 protocol. So, I think it may stay in the image as is and the Mutliboot2 header should not cover .reloc section.

By the way, why do we need .reloc section in the PE image? Is not %rip relative addressing sufficient? IIRC the Linux kernel just contains a stub .reloc section. Could not we do the same?

Quoted 3 lines

compressed linker output, but a post-processed (by mkelf32) image, maybe what we need is a build tool doing similar post-processing on xen.efi? Otoh getting disk image and in-memory image aligned ought

Yep, this should work too.

Quoted 2 lines

to be possible by setting --section-alignment= and --file-alignment= to the same value (resulting in a much larger file) - adjusting file

IIRC this did not work for some reason. Maybe it would be better to enforce correct alignment and required padding using linker script.

Quoted 7 lines

positions would effectively be what a post-processing tool would need to do (like with mkelf32 perhaps we could then at least save the first ~2Mb of space). Which would still leave .reloc to be dealt with
- maybe we could place this after .init, but still ahead of __init_end (such that the memory would get freed late in the boot process). Not sure whether EFI loaders would "like" such an unusual placement.

Yeah, good question...

Quoted 2 lines

Also not sure what to do with Dwarf debug info, which just recently we managed to avoid needing to strip unconditionally.

I think debug info may stay as is. Just Multiboot2 header should not cover it if it is not needed.

Quoted 8 lines
Quoted 8 lines
Quoted 7 lines
Quoted 3 lines

Maybe
newer binutils are more flexible and will be able to produce a PE image with properties required by Multiboot2 protocol.

Isn't all you need the MB2 header within the first so many bytes of the (disk) image? Or was it the image as loaded into memory? Both should be possible to arrange for.

Quoted 8 lines
Quoted 2 lines

IIRC Multiboot2 protocol requires the header in first 32 kiB of an image. So, this is not a problem.

I was about to ask "Disk image or in-memory image?" But this won't matter if the image as a whole got linearized, as long as the first section doesn't start to high up. I notice that xen-syms doesn't fit this requirement either, only the output of mkelf32 does. Which suggests that there may not be a way around a post-processing tool.

Could not we drop 2 MiB padding at the beginning of xen-syms by changing some things in the linker script?

Daniel

#14 lore raw

Re: [PATCH v3 2/5] xen/x86: manually build xen.mb.efi binary

Quoted 34 lines · Daniel Kiper
Quoted 29 lines · Jan Beulich
Quoted 17 lines · Daniel Kiper
Quoted 7 lines · Jan Beulich
Quoted 6 lines · Daniel Kiper
Quoted 5 lines · Jan Beulich
Quoted 3 lines · Bob Eshleman

What is your intuition WRT the idea that instead of trying add a PE/COFF hdr in front of Xen's mb2 bin, we instead go the route of introducing valid mb2 entry points into xen.efi?

Quoted 8 lines
Quoted 6 lines
Quoted 2 lines

At the first glance I think this is going to be less intrusive, and hence to be preferred. But of course I haven't experimented in any way ...

When I worked on this a few years ago I tried that way. Sadly I failed because I was not able to produce "linear" PE image using binutils exiting that days.

What is a "linear" PE image?

Quoted 10 lines

The problem with Multiboot family protocols is that all code and data sections have to be glued together in the image and as such loaded into the memory (IIRC BSS is an exception but it has to live behind the image). So, you cannot use PE image which has different representation in file and memory. IIRC by default at least code and data sections in xen.efi have different sizes in PE file and in memory. I tried to fix that using linker script and objcopy but it did not work. Sadly I do not remember the details but there is pretty good chance you can find relevant emails in Xen-devel archive with me explaining what kind of problems I met.

Quoted 3 lines

Ah, this rings a bell. Even the .bss-is-last assumption doesn't hold, because .reloc (for us as well as in general) comes later, but needs loading (in the right place). Since even xen.gz isn't simply the

Quoted 7 lines

However, IIRC it is not used when Xen is loaded through Multiboot2 protocol. So, I think it may stay in the image as is and the Mutliboot2 header should not cover .reloc section.

By the way, why do we need .reloc section in the PE image? Is not %rip relative addressing sufficient? IIRC the Linux kernel just contains a stub .reloc section. Could not we do the same?

%rip-relative addressing can (obviously, I think) help only for text. But we also have data containing pointers, which need relocating.

Quoted 11 lines
Quoted 3 lines

compressed linker output, but a post-processed (by mkelf32) image, maybe what we need is a build tool doing similar post-processing on xen.efi? Otoh getting disk image and in-memory image aligned ought

Yep, this should work too.

Quoted 2 lines

to be possible by setting --section-alignment= and --file-alignment= to the same value (resulting in a much larger file) - adjusting file

IIRC this did not work for some reason. Maybe it would be better to enforce correct alignment and required padding using linker script.

I'm not convinced the linker script is the correct vehicle here. It is mainly about placement in the address space (i.e. laying out how things will end up in memory), not about file layout.

Quoted 8 lines
Quoted 7 lines

positions would effectively be what a post-processing tool would need to do (like with mkelf32 perhaps we could then at least save the first ~2Mb of space). Which would still leave .reloc to be dealt with
- maybe we could place this after .init, but still ahead of __init_end (such that the memory would get freed late in the boot process). Not sure whether EFI loaders would "like" such an unusual placement.

Quoted 7 lines

Yeah, good question...

Quoted 2 lines

Also not sure what to do with Dwarf debug info, which just recently we managed to avoid needing to strip unconditionally.

I think debug info may stay as is. Just Multiboot2 header should not cover it if it is not needed.

You did say that .bss is expected to be last, which both .reloc and debug info violate.

Quoted 11 lines
Quoted 11 lines
Quoted 10 lines
Quoted 7 lines
Quoted 3 lines

Maybe
newer binutils are more flexible and will be able to produce a PE image with properties required by Multiboot2 protocol.

Isn't all you need the MB2 header within the first so many bytes of the (disk) image? Or was it the image as loaded into memory? Both should be possible to arrange for.

IIRC Multiboot2 protocol requires the header in first 32 kiB of an image. So, this is not a problem.

Quoted 8 lines
Quoted 5 lines

I was about to ask "Disk image or in-memory image?" But this won't matter if the image as a whole got linearized, as long as the first section doesn't start to high up. I notice that xen-syms doesn't fit this requirement either, only the output of mkelf32 does. Which suggests that there may not be a way around a post-processing tool.

Could not we drop 2 MiB padding at the beginning of xen-syms by changing some things in the linker script?

Not sure, but at the first glance I don't think so. If we want section and file alignment to match, and if we want sections at 2Mb boundaries, then the first section - since it cannot start at 0 - will need to be at 2Mb. In xen-syms the linker manages to put it at 32kb, but I think arranging for such also for PE output (despite both alignments set to match) would require a linker change, perhaps even a new command line option (because the two alignment requests can't be silently ignored).

Jan

#15 lore raw

Re: [PATCH v3 2/5] xen/x86: manually build xen.mb.efi binary

Quoted 39 lines · Jan Beulich
Quoted 38 lines · Daniel Kiper
Quoted 29 lines · Jan Beulich
Quoted 17 lines · Daniel Kiper
Quoted 7 lines · Jan Beulich
Quoted 6 lines · Daniel Kiper
Quoted 5 lines · Jan Beulich
Quoted 3 lines · Bob Eshleman

What is your intuition WRT the idea that instead of trying add a PE/COFF hdr in front of Xen's mb2 bin, we instead go the route of introducing valid mb2 entry points into xen.efi?

Quoted 8 lines
Quoted 6 lines
Quoted 2 lines

At the first glance I think this is going to be less intrusive, and hence to be preferred. But of course I haven't experimented in any way ...

When I worked on this a few years ago I tried that way. Sadly I failed because I was not able to produce "linear" PE image using binutils exiting that days.

What is a "linear" PE image?

Quoted 10 lines

The problem with Multiboot family protocols is that all code and data sections have to be glued together in the image and as such loaded into the memory (IIRC BSS is an exception but it has to live behind the image). So, you cannot use PE image which has different representation in file and memory. IIRC by default at least code and data sections in xen.efi have different sizes in PE file and in memory. I tried to fix that using linker script and objcopy but it did not work. Sadly I do not remember the details but there is pretty good chance you can find relevant emails in Xen-devel archive with me explaining what kind of problems I met.

Quoted 3 lines

Ah, this rings a bell. Even the .bss-is-last assumption doesn't hold, because .reloc (for us as well as in general) comes later, but needs loading (in the right place). Since even xen.gz isn't simply the

However, IIRC it is not used when Xen is loaded through Multiboot2 protocol. So, I think it may stay in the image as is and the Mutliboot2 header should not cover .reloc section.

Quoted 6 lines
Quoted 3 lines

By the way, why do we need .reloc section in the PE image? Is not %rip relative addressing sufficient? IIRC the Linux kernel just contains a stub .reloc section. Could not we do the same?

%rip-relative addressing can (obviously, I think) help only for text. But we also have data containing pointers, which need relocating.

Ahhh, right, I totally forgot about it.

Quoted 9 lines
Quoted 9 lines
Quoted 3 lines

compressed linker output, but a post-processed (by mkelf32) image, maybe what we need is a build tool doing similar post-processing on xen.efi? Otoh getting disk image and in-memory image aligned ought

Yep, this should work too.

Quoted 2 lines

to be possible by setting --section-alignment= and --file-alignment= to the same value (resulting in a much larger file) - adjusting file

Quoted 6 lines
Quoted 2 lines

IIRC this did not work for some reason. Maybe it would be better to enforce correct alignment and required padding using linker script.

I'm not convinced the linker script is the correct vehicle here. It is mainly about placement in the address space (i.e. laying out how things will end up in memory), not about file layout.

OK but at least I would check what is possible and do it then.

Quoted 10 lines
Quoted 10 lines
Quoted 7 lines

positions would effectively be what a post-processing tool would need to do (like with mkelf32 perhaps we could then at least save the first ~2Mb of space). Which would still leave .reloc to be dealt with
- maybe we could place this after .init, but still ahead of __init_end (such that the memory would get freed late in the boot process). Not sure whether EFI loaders would "like" such an unusual placement.

Yeah, good question...

Quoted 8 lines
Quoted 5 lines
Quoted 2 lines

Also not sure what to do with Dwarf debug info, which just recently we managed to avoid needing to strip unconditionally.

I think debug info may stay as is. Just Multiboot2 header should not cover it if it is not needed.

You did say that .bss is expected to be last, which both .reloc and debug info violate.

The .bss section has to be last one in memory from Multiboot2 protocol point of view. However, nothing, AFAICT, forbids to have something behind in the file. Of course if you ignore the data at the end of file when you load the image using Multiboot2 protocol.

Daniel

#16 lore raw

Re: [PATCH v3 2/5] xen/x86: manually build xen.mb.efi binary

Quoted 9 lines · Daniel Kiper
Quoted 8 lines · Jan Beulich
Quoted 6 lines · Daniel Kiper
Quoted 2 lines · Jan Beulich

Also not sure what to do with Dwarf debug info, which just recently we managed to avoid needing to strip unconditionally.

I think debug info may stay as is. Just Multiboot2 header should not cover it if it is not needed.

Quoted 7 lines
Quoted 2 lines

You did say that .bss is expected to be last, which both .reloc and debug info violate.

The .bss section has to be last one in memory from Multiboot2 protocol point of view. However, nothing, AFAICT, forbids to have something behind in the file. Of course if you ignore the data at the end of file when you load the image using Multiboot2 protocol.

Well, debug info can be ignored. If MB2 would work like it does today, then .reloc also would never be touched. Feels a little fragile, but might be okay then.

Jan

#17 lore raw

PATCH v3 3/5 xen/x86: add some addresses to the Multiboot header

In reply to Bobby Eshleman

From: Daniel Kiper <[email protected]>

In comparison to ELF the PE format is not supported by the Multiboot protocol. So, if we wish to load xen.mb.efi using this protocol we have to put header_addr, load_addr, load_end_addr, bss_end_addr and entry_addr data into Multiboot header.

The Multiboot protocol spec can be found at
https://www.gnu.org/software/grub/manual/multiboot/

---
 xen/arch/x86/boot/head.S | 13 ++++++++++++-
 1 file changed, 12 insertions(+), 1 deletion(-)
xen/arch/x86/boot/head.S +12 -1
index 2987b4f03a..189d91a872 100644
@@ -50,13 +50,24 @@ ENTRY(start)
         .balign 4
 multiboot1_header:             /*** MULTIBOOT1 HEADER ****/
 #define MULTIBOOT_HEADER_FLAGS (MULTIBOOT_HEADER_MODS_ALIGNED | \
-                                MULTIBOOT_HEADER_WANT_MEMORY)
+                                MULTIBOOT_HEADER_WANT_MEMORY | \
+                                MULTIBOOT_HEADER_HAS_ADDR)
         /* Magic number indicating a Multiboot header. */
         .long   MULTIBOOT_HEADER_MAGIC
         /* Flags to bootloader (see Multiboot spec). */
         .long   MULTIBOOT_HEADER_FLAGS
         /* Checksum: must be the negated sum of the first two fields. */
         .long   -(MULTIBOOT_HEADER_MAGIC + MULTIBOOT_HEADER_FLAGS)
+        /* header_addr */
+        .long   sym_offs(multiboot1_header)
+        /* load_addr */
+        .long   sym_offs(start)
+        /* load_end_addr */
+        .long   sym_offs(__bss_start)
+        /* bss_end_addr */
+        .long   sym_offs(__2M_rwdata_end)
+        /* entry_addr */
+        .long   sym_offs(__start)
 
         .size multiboot1_header, . - multiboot1_header
         .type multiboot1_header, @object
Signature · 1 line
2.30.0
#18 lore raw

Re: [PATCH v3 3/5] xen/x86: add some addresses to the Multiboot header

Quoted 9 lines · Bobby Eshleman

From: Daniel Kiper <[email protected]>

In comparison to ELF the PE format is not supported by the Multiboot protocol. So, if we wish to load xen.mb.efi using this protocol we have to put header_addr, load_addr, load_end_addr, bss_end_addr and entry_addr data into Multiboot header.

The Multiboot protocol spec can be found at
https://www.gnu.org/software/grub/manual/multiboot/

And because of this spec saying "the boot loader should use them instead of the fields in the actual executable header to calculate where to load the OS image" this change will affect the ELF image as well. For example ...

Quoted 13 lines
xen/arch/x86/boot/head.S +2 -1
@@ -50,13 +50,24 @@ ENTRY(start)
         .balign 4
 multiboot1_header:             /*** MULTIBOOT1 HEADER ****/
 #define MULTIBOOT_HEADER_FLAGS (MULTIBOOT_HEADER_MODS_ALIGNED | \
-                                MULTIBOOT_HEADER_WANT_MEMORY)
+                                MULTIBOOT_HEADER_WANT_MEMORY | \
+                                MULTIBOOT_HEADER_HAS_ADDR)
         /* Magic number indicating a Multiboot header. */
         .long   MULTIBOOT_HEADER_MAGIC
         /* Flags to bootloader (see Multiboot spec). */
         .long   MULTIBOOT_HEADER_FLAGS

xen/arch/x86/boot/head.S ENTRY(start) · line 68

Quoted 10 lines
xen/arch/x86/boot/head.S +8 -0
         /* Checksum: must be the negated sum of the first two fields. */
         .long   -(MULTIBOOT_HEADER_MAGIC + MULTIBOOT_HEADER_FLAGS)
+        /* header_addr */
+        .long   sym_offs(multiboot1_header)
+        /* load_addr */
+        .long   sym_offs(start)
+        /* load_end_addr */
+        .long   sym_offs(__bss_start)
+        /* bss_end_addr */
+        .long   sym_offs(__2M_rwdata_end)

... the ELF image end at _end, not at __2M_rwdata_end. I realize that with 2M alignment in use, this may actually be a problem, as one of the modules (the initrd in particular) could be placed overlapping the (_end, __2M_rwdata_end) range. Nevertheless I think you want to specify _end (or __bss_end) here.

As to the initial point made - would it be possible to leave the flag unset in the EFL image and force it set only in xen.mb.efi? Yes, this may require yet another post-processing step.

Jan

Quoted 6 lines
xen/arch/x86/boot/head.S +2 -0
+        /* entry_addr */
+        .long   sym_offs(__start)
 
         .size multiboot1_header, . - multiboot1_header
         .type multiboot1_header, @object
#19 lore raw

PATCH v3 4/5 xen/x86: add some addresses to the Multiboot2 header

In reply to Bobby Eshleman

From: Daniel Kiper <[email protected]>

In comparison to ELF the PE format is not supported by the Multiboot2 protocol. So, if we wish to load xen.mb.efi using this protocol we have to add MULTIBOOT2_HEADER_TAG_ADDRESS and MULTIBOOT2_HEADER_TAG_ENTRY_ADDRESS tags into Multiboot2 header.

Additionally, put MULTIBOOT2_HEADER_TAG_ENTRY_ADDRESS and MULTIBOOT2_HEADER_TAG_ENTRY_ADDRESS_EFI64 tags close to each other to make the header more readable.

The Multiboot2 protocol spec can be found at
https://www.gnu.org/software/grub/manual/multiboot2/

---
 xen/arch/x86/boot/head.S | 19 +++++++++++++++----
 1 file changed, 15 insertions(+), 4 deletions(-)
xen/arch/x86/boot/head.S +15 -4
index 189d91a872..f2edd182a5 100644
@@ -94,6 +94,13 @@ multiboot2_header:
         /* Align modules at page boundry. */
         mb2ht_init MB2_HT(MODULE_ALIGN), MB2_HT(REQUIRED)
 
+        /* The address tag. */
+        mb2ht_init MB2_HT(ADDRESS), MB2_HT(REQUIRED), \
+                   sym_offs(multiboot2_header), /* header_addr */ \
+                   sym_offs(start),             /* load_addr */ \
+                   sym_offs(__bss_start),       /* load_end_addr */ \
+                   sym_offs(__2M_rwdata_end)    /* bss_end_addr */
+
         /* Load address preference. */
         mb2ht_init MB2_HT(RELOCATABLE), MB2_HT(OPTIONAL), \
                    sym_offs(start), /* Min load address. */ \
@@ -101,6 +108,14 @@ multiboot2_header:
                    0x200000, /* Load address alignment (2 MiB). */ \
                    MULTIBOOT2_LOAD_PREFERENCE_HIGH
 
+        /* Multiboot2 entry point. */
+        mb2ht_init MB2_HT(ENTRY_ADDRESS), MB2_HT(REQUIRED), \
+                   sym_offs(__start)
+
+        /* EFI64 Multiboot2 entry point. */
+        mb2ht_init MB2_HT(ENTRY_ADDRESS_EFI64), MB2_HT(OPTIONAL), \
+                   sym_offs(__efi64_mb2_start)
+
         /* Console flags tag. */
         mb2ht_init MB2_HT(CONSOLE_FLAGS), MB2_HT(OPTIONAL), \
                    MULTIBOOT2_CONSOLE_FLAGS_EGA_TEXT_SUPPORTED
@@ -114,10 +129,6 @@ multiboot2_header:
         /* Request that ExitBootServices() not be called. */
         mb2ht_init MB2_HT(EFI_BS), MB2_HT(OPTIONAL)
 
-        /* EFI64 Multiboot2 entry point. */
-        mb2ht_init MB2_HT(ENTRY_ADDRESS_EFI64), MB2_HT(OPTIONAL), \
-                   sym_offs(__efi64_mb2_start)
-
         /* Multiboot2 header end tag. */
         mb2ht_init MB2_HT(END), MB2_HT(REQUIRED)
 .Lmultiboot2_header_end:
Signature · 1 line
2.30.0
#20 lore raw

Re: [PATCH v3 4/5] xen/x86: add some addresses to the Multiboot2 header

Quoted 25 lines · Bobby Eshleman

From: Daniel Kiper <[email protected]>

In comparison to ELF the PE format is not supported by the Multiboot2 protocol. So, if we wish to load xen.mb.efi using this protocol we have to add MULTIBOOT2_HEADER_TAG_ADDRESS and MULTIBOOT2_HEADER_TAG_ENTRY_ADDRESS tags into Multiboot2 header.

Additionally, put MULTIBOOT2_HEADER_TAG_ENTRY_ADDRESS and MULTIBOOT2_HEADER_TAG_ENTRY_ADDRESS_EFI64 tags close to each other to make the header more readable.

The Multiboot2 protocol spec can be found at
https://www.gnu.org/software/grub/manual/multiboot2/

---
 xen/arch/x86/boot/head.S | 19 +++++++++++++++----
 1 file changed, 15 insertions(+), 4 deletions(-)
xen/arch/x86/boot/head.S +0 -0
index 189d91a872..f2edd182a5 100644

xen/arch/x86/boot/head.S multiboot2_header: · line 102

Quoted 9 lines
xen/arch/x86/boot/head.S +6 -0
@@ -94,6 +94,13 @@ multiboot2_header:
         /* Align modules at page boundry. */
         mb2ht_init MB2_HT(MODULE_ALIGN), MB2_HT(REQUIRED)
 
+        /* The address tag. */
+        mb2ht_init MB2_HT(ADDRESS), MB2_HT(REQUIRED), \
+                   sym_offs(multiboot2_header), /* header_addr */ \
+                   sym_offs(start),             /* load_addr */ \
+                   sym_offs(__bss_start),       /* load_end_addr */ \
+                   sym_offs(__2M_rwdata_end)    /* bss_end_addr */

Shouldn't this only be present when a PE binary is built?

You seem to unconditionally add this to the header, even when the resulting binary will be in ELF format?

According to the spec: "This information does not need to be provided if the kernel image is in ELF format", and hence Xen shouldn't require the loader to understand this tag unless it's strictly required, as the presence of the tag forces the bootloader to use the presented information in order to load the kernel, regardless of the underlying binary format.

Thanks, Roger.

#21 lore raw

Re: [PATCH v3 4/5] xen/x86: add some addresses to the Multiboot2 header

Quoted 41 lines · Roger Pau Monné
Quoted 25 lines · Bobby Eshleman

From: Daniel Kiper <[email protected]>

In comparison to ELF the PE format is not supported by the Multiboot2 protocol. So, if we wish to load xen.mb.efi using this protocol we have to add MULTIBOOT2_HEADER_TAG_ADDRESS and MULTIBOOT2_HEADER_TAG_ENTRY_ADDRESS tags into Multiboot2 header.

Additionally, put MULTIBOOT2_HEADER_TAG_ENTRY_ADDRESS and MULTIBOOT2_HEADER_TAG_ENTRY_ADDRESS_EFI64 tags close to each other to make the header more readable.

The Multiboot2 protocol spec can be found at
https://www.gnu.org/software/grub/manual/multiboot2/

---
 xen/arch/x86/boot/head.S | 19 +++++++++++++++----
 1 file changed, 15 insertions(+), 4 deletions(-)
xen/arch/x86/boot/head.S +0 -0
index 189d91a872..f2edd182a5 100644

xen/arch/x86/boot/head.S multiboot2_header: · line 102

Quoted 9 lines
xen/arch/x86/boot/head.S +6 -0
@@ -94,6 +94,13 @@ multiboot2_header:
         /* Align modules at page boundry. */
         mb2ht_init MB2_HT(MODULE_ALIGN), MB2_HT(REQUIRED)
 
+        /* The address tag. */
+        mb2ht_init MB2_HT(ADDRESS), MB2_HT(REQUIRED), \
+                   sym_offs(multiboot2_header), /* header_addr */ \
+                   sym_offs(start),             /* load_addr */ \
+                   sym_offs(__bss_start),       /* load_end_addr */ \
+                   sym_offs(__2M_rwdata_end)    /* bss_end_addr */

Shouldn't this only be present when a PE binary is built?

You seem to unconditionally add this to the header, even when the resulting binary will be in ELF format?

Quoted 9 lines

According to the spec: "This information does not need to be provided if the kernel image is in ELF format", and hence Xen shouldn't require the loader to understand this tag unless it's strictly required, as the presence of the tag forces the bootloader to use the presented information in order to load the kernel, regardless of the underlying binary format.

Thanks, Roger.

Ah yes, this is true. It may have made more sense to do this with v2 trying to step us in the direction of a single unified binary, but it certainly isn't required with v3.

Thanks,
Bob

#22 lore raw

PATCH v3 5/5 xen/x86/efi: Verify dom0 kernel with SHIM_LOCK protocol in efi_multiboot2()

In reply to Bobby Eshleman

From: Daniel Kiper <[email protected]>

This splits out efi_shim_lock() into common code and uses it to verify the dom0 kernel in efi_multiboot2().

---
 xen/arch/x86/boot/head.S    | 20 ++++++++++++++++++--
 xen/arch/x86/efi/efi-boot.h |  6 ++++++
 xen/arch/x86/efi/stub.c     |  5 ++++-
 xen/common/efi/boot.c       | 19 +++++++++++++------
 4 files changed, 41 insertions(+), 9 deletions(-)
xen/arch/x86/boot/head.S +18 -2
index f2edd182a5..943792eb43 100644
@@ -244,9 +244,13 @@ __efi64_mb2_start:
         jmp     x86_32_switch
 
 .Lefi_multiboot2_proto:
-        /* Zero EFI SystemTable and EFI ImageHandle addresses. */
+        /*
+         * Zero EFI SystemTable, EFI ImageHandle and
+         * dom0 kernel module struct addresses.
+         */
         xor     %esi,%esi
         xor     %edi,%edi
+        xor     %r14d, %r14d
 
         /* Skip Multiboot2 information fixed part. */
         lea     (MB2_fixed_sizeof+MULTIBOOT2_TAG_ALIGN-1)(%rbx),%ecx
@@ -284,6 +288,15 @@ __efi64_mb2_start:
         cmove   MB2_efi64_ih(%rcx),%rdi
         je      .Lefi_mb2_next_tag
 
+        /* Get Dom0 kernel module struct address from Multiboot2 information. */
+        cmpl    $MULTIBOOT2_TAG_TYPE_MODULE,MB2_tag_type(%rcx)
+        jne     .Lefi_mb2_end
+
+        test    %r14d, %r14d
+        cmovz   %ecx, %r14d
+        jmp     .Lefi_mb2_next_tag
+
+.Lefi_mb2_end:
         /* Is it the end of Multiboot2 information? */
         cmpl    $MULTIBOOT2_TAG_TYPE_END,MB2_tag_type(%rcx)
         je      .Lrun_bs
@@ -345,9 +358,12 @@ __efi64_mb2_start:
         /* Keep the stack aligned. Do not pop a single item off it. */
         mov     (%rsp),%rdi
 
+        mov     %r14d, %edx
+
         /*
          * efi_multiboot2() is called according to System V AMD64 ABI:
-         *   - IN:  %rdi - EFI ImageHandle, %rsi - EFI SystemTable.
+         *   - IN: %rdi - EFI ImageHandle, %rsi - EFI SystemTable,
+         *         %rdx - Dom0 kernel module struct address.
          */
         call    efi_multiboot2
 
xen/arch/x86/efi/efi-boot.h +6 -0
index f694a069c9..0d025ad9a5 100644
@@ -3,6 +3,8 @@
  * is intended to be included by common/efi/boot.c _only_, and
  * therefore can define arch specific global variables.
  */
+#include <xen/types.h>
+#include <xen/multiboot2.h>
 #include <xen/vga.h>
 #include <asm/e820.h>
 #include <asm/edd.h>
@@ -762,6 +764,10 @@ void __init efi_multiboot2(EFI_HANDLE ImageHandle,
 
     gop = efi_get_gop();
 
+    if ( dom0_kernel && dom0_kernel->mod_end > dom0_kernel->mod_start )
+        efi_shim_lock((VOID *)(unsigned long)dom0_kernel->mod_start,
+                      dom0_kernel->mod_end - dom0_kernel->mod_start);
+
     if ( gop )
         gop_mode = efi_find_gop_mode(gop, 0, 0, 0);
 
xen/arch/x86/efi/stub.c +4 -1
index 9bd6355ec3..7d459905fa 100644
@@ -1,7 +1,9 @@
+#include <xen/types.h>
 #include <xen/efi.h>
 #include <xen/errno.h>
 #include <xen/init.h>
 #include <xen/lib.h>
+#include <xen/multiboot2.h>
 #include <asm/asm_defns.h>
 #include <asm/efibind.h>
 #include <asm/page.h>
@@ -29,7 +31,8 @@ asm (
     );
 
 void __init noreturn efi_multiboot2(EFI_HANDLE ImageHandle,
-                                    EFI_SYSTEM_TABLE *SystemTable)
+                                    EFI_SYSTEM_TABLE *SystemTable,
+                                    multiboot2_tag_module_t *dom0_kernel)
 {
     static const CHAR16 __initconst err[] =
         L"Xen does not have EFI code build in!\r\nSystem halted!\r\n";
xen/common/efi/boot.c +13 -6
index 63e289ab85..8ce6715b59 100644
@@ -133,6 +133,7 @@ static void efi_console_set_mode(void);
 static EFI_GRAPHICS_OUTPUT_PROTOCOL *efi_get_gop(void);
 static UINTN efi_find_gop_mode(EFI_GRAPHICS_OUTPUT_PROTOCOL *gop,
                                UINTN cols, UINTN rows, UINTN depth);
+static void efi_shim_lock(const VOID *Buffer, UINT32 Size);
 static void efi_tables(void);
 static void setup_efi_pci(void);
 static void efi_variables(void);
@@ -830,6 +831,17 @@ static UINTN __init efi_find_gop_mode(EFI_GRAPHICS_OUTPUT_PROTOCOL *gop,
     return gop_mode;
 }
 
+static void __init efi_shim_lock(const VOID *Buffer, UINT32 Size)
+{
+    static EFI_GUID __initdata shim_lock_guid = SHIM_LOCK_PROTOCOL_GUID;
+    EFI_SHIM_LOCK_PROTOCOL *shim_lock;
+    EFI_STATUS status;
+
+    if ( !EFI_ERROR(efi_bs->LocateProtocol(&shim_lock_guid, NULL, (void **)&shim_lock)) &&
+         (status = shim_lock->Verify(Buffer, Size)) != EFI_SUCCESS )
+        PrintErrMesg(L"Dom0 kernel image could not be verified", status);
+}
+
 static void __init efi_tables(void)
 {
     unsigned int i;
@@ -1123,13 +1135,11 @@ void EFIAPI __init noreturn
 efi_start(EFI_HANDLE ImageHandle, EFI_SYSTEM_TABLE *SystemTable)
 {
     static EFI_GUID __initdata loaded_image_guid = LOADED_IMAGE_PROTOCOL;
-    static EFI_GUID __initdata shim_lock_guid = SHIM_LOCK_PROTOCOL_GUID;
     EFI_LOADED_IMAGE *loaded_image;
     EFI_STATUS status;
     unsigned int i, argc;
     CHAR16 **argv, *file_name, *cfg_file_name = NULL, *options = NULL;
     UINTN gop_mode = ~0;
-    EFI_SHIM_LOCK_PROTOCOL *shim_lock;
     EFI_GRAPHICS_OUTPUT_PROTOCOL *gop = NULL;
     union string section = { NULL }, name;
     bool base_video = false;
@@ -1296,10 +1306,7 @@ efi_start(EFI_HANDLE ImageHandle, EFI_SYSTEM_TABLE *SystemTable)
             read_file(dir_handle, s2w(&name), &kernel, option_str);
             efi_bs->FreePool(name.w);
 
-            if ( !EFI_ERROR(efi_bs->LocateProtocol(&shim_lock_guid, NULL,
-                            (void **)&shim_lock)) &&
-                 (status = shim_lock->Verify(kernel.ptr, kernel.size)) != EFI_SUCCESS )
-                PrintErrMesg(L"Dom0 kernel image could not be verified", status);
+            efi_shim_lock(kernel.ptr, kernel.size);
         }
 
         if ( !read_section(loaded_image, L"ramdisk", &ramdisk, NULL) )
Signature · 1 line
2.30.0
#23 lore raw

Re: [PATCH v3 5/5] xen/x86/efi: Verify dom0 kernel with SHIM_LOCK protocol in efi_multiboot2()

Quoted 4 lines · Bobby Eshleman
xen/arch/x86/boot/head.S +0 -0
@@ -244,9 +244,13 @@ __efi64_mb2_start:
         jmp     x86_32_switch

xen/arch/x86/boot/head.S __efi64_mb2_start: · line 253

Quoted 10 lines
xen/arch/x86/boot/head.S +5 -1
 
 .Lefi_multiboot2_proto:
-        /* Zero EFI SystemTable and EFI ImageHandle addresses. */
+        /*
+         * Zero EFI SystemTable, EFI ImageHandle and
+         * dom0 kernel module struct addresses.
+         */
         xor     %esi,%esi
         xor     %edi,%edi
+        xor     %r14d, %r14d

Nit: There's little point in having the d suffixes here and below, and the code would be slightly easier to read without.

xen/arch/x86/boot/head.S __efi64_mb2_start: · line 292

Quoted 8 lines
xen/arch/x86/boot/head.S +2 -0
         /* Skip Multiboot2 information fixed part. */
         lea     (MB2_fixed_sizeof+MULTIBOOT2_TAG_ALIGN-1)(%rbx),%ecx
@@ -284,6 +288,15 @@ __efi64_mb2_start:
         cmove   MB2_efi64_ih(%rcx),%rdi
         je      .Lefi_mb2_next_tag
 
+        /* Get Dom0 kernel module struct address from Multiboot2 information. */
+        cmpl    $MULTIBOOT2_TAG_TYPE_MODULE,MB2_tag_type(%rcx)

Not: If elsewhere in the code additions you put blanks after the comma (which I appreciate), please do so here as well.

xen/arch/x86/boot/head.S __efi64_mb2_start: · line 296

Quoted 4 lines
xen/arch/x86/boot/head.S +4 -0
+        jne     .Lefi_mb2_end
+
+        test    %r14d, %r14d
+        cmovz   %ecx, %r14d

So this doesn't truncate the address because higher up %ecx was loaded instead of %rcx. I realize that's not code you add, but it still strikes me as odd. Are there indeed guarantees that all of this will live below 4Gb?

xen/arch/x86/efi/efi-boot.h void __init efi_multiboot2(EFI_HANDLE ImageHandle,

Quoted 12 lines
xen/arch/x86/efi/efi-boot.h +2 -0
@@ -3,6 +3,8 @@
  * is intended to be included by common/efi/boot.c _only_, and
  * therefore can define arch specific global variables.
  */
+#include <xen/types.h>
+#include <xen/multiboot2.h>
 #include <xen/vga.h>
 #include <asm/e820.h>
 #include <asm/edd.h>
@@ -762,6 +764,10 @@ void __init efi_multiboot2(EFI_HANDLE ImageHandle,

Isn't there a hunk missing up from here to add the new parameter to efi_multiboot2()?

xen/arch/x86/efi/efi-boot.h void __init efi_multiboot2(EFI_HANDLE ImageHandle, · line 769

Quoted 5 lines
xen/arch/x86/efi/efi-boot.h +3 -0
     gop = efi_get_gop();
 
+    if ( dom0_kernel && dom0_kernel->mod_end > dom0_kernel->mod_start )
+        efi_shim_lock((VOID *)(unsigned long)dom0_kernel->mod_start,
+                      dom0_kernel->mod_end - dom0_kernel->mod_start);

While somewhat unrelated to the change itself - how come the fields are all u32 (and hence you need to cast to unsigned long first)? There having been requests to allow for about 1Gb initrd images, I find it quite reasonable to expect that modules may not all fit below 4Gb.

xen/arch/x86/efi/stub.c line 1

Quoted 4 lines
xen/arch/x86/efi/stub.c +1 -0
@@ -1,7 +1,9 @@
+#include <xen/types.h>

Please don't, even less so without honoring the alphabetical sorting.

Quoted 9 lines
xen/arch/x86/efi/stub.c +1 -0
 #include <xen/efi.h>
 #include <xen/errno.h>
 #include <xen/init.h>
 #include <xen/lib.h>
+#include <xen/multiboot2.h>
 #include <asm/asm_defns.h>
 #include <asm/efibind.h>
 #include <asm/page.h>

xen/arch/x86/efi/stub.c asm ( · line 35

Quoted 6 lines
xen/arch/x86/efi/stub.c +2 -1
@@ -29,7 +31,8 @@ asm (
     );
 
 void __init noreturn efi_multiboot2(EFI_HANDLE ImageHandle,
-                                    EFI_SYSTEM_TABLE *SystemTable)
+                                    EFI_SYSTEM_TABLE *SystemTable,
+                                    multiboot2_tag_module_t *dom0_kernel)

const?

Quoted 11 lines
xen/common/efi/boot.c +1 -0
@@ -133,6 +133,7 @@ static void efi_console_set_mode(void);
 static EFI_GRAPHICS_OUTPUT_PROTOCOL *efi_get_gop(void);
 static UINTN efi_find_gop_mode(EFI_GRAPHICS_OUTPUT_PROTOCOL *gop,
                                UINTN cols, UINTN rows, UINTN depth);
+static void efi_shim_lock(const VOID *Buffer, UINT32 Size);
 static void efi_tables(void);
 static void setup_efi_pci(void);
 static void efi_variables(void);

xen/common/efi/boot.c static UINTN __init efi_find_gop_mode(EFI_GRAPHICS_OUTPUT_PROTOCOL *gop, · line 834

Quoted 4 lines
xen/common/efi/boot.c +1 -0
@@ -830,6 +831,17 @@ static UINTN __init efi_find_gop_mode(EFI_GRAPHICS_OUTPUT_PROTOCOL *gop,
     return gop_mode;
 }
 
+static void __init efi_shim_lock(const VOID *Buffer, UINT32 Size)

Maybe better efi_shim_lock_verify()?

xen/common/efi/boot.c static UINTN __init efi_find_gop_mode(EFI_GRAPHICS_OUTPUT_PROTOCOL *gop, · line 840

Quoted 6 lines
xen/common/efi/boot.c +6 -0
+{
+    static EFI_GUID __initdata shim_lock_guid = SHIM_LOCK_PROTOCOL_GUID;
+    EFI_SHIM_LOCK_PROTOCOL *shim_lock;
+    EFI_STATUS status;
+
+    if ( !EFI_ERROR(efi_bs->LocateProtocol(&shim_lock_guid, NULL, (void **)&shim_lock)) &&

Nit: Overly long line.

xen/common/efi/boot.c static UINTN __init efi_find_gop_mode(EFI_GRAPHICS_OUTPUT_PROTOCOL *gop, · line 842

Quoted 2 lines
xen/common/efi/boot.c +2 -0
+         (status = shim_lock->Verify(Buffer, Size)) != EFI_SUCCESS )
+        PrintErrMesg(L"Dom0 kernel image could not be verified", status);

I'm willing to let it be as is, but in principle this function is not Dom0-specific the way you've split it out. _If_ you leave it this way, perhaps (on top of the suggestion above) perhaps better name it efi_shim_lock_verify_dom0()?

Jan

#24 lore raw
In reply to Bobby Eshleman
Quoted 7 lines · Bobby Eshleman

This is version 3 for a patch set sent out to the ML in 2018 [1] to support UEFI Secure Boot for Xen on multiboot2 platforms.

A new binary, xen.mb.efi, is built. It contains the mb2 header as well as a hand-crafted PE/COFF header. The dom0 kernel is verified using the shim lock protocol.

Quoted 6 lines

I followed with v2 feedback and attempted to convert the PE/COFF header into C instead of ASM. Unfortunately, this was only possible for the first part (Legacy) of the PE/COFF header. The other parts required addresses only available at link time (such as __2M_rwdata_end, __pe_SizeOfImage, efi_mb_start address, etc...), which effectively ruled out C.

I don't follow the conclusion drawn, so would you mind going into further detail?

Jan

#25 lore raw
Quoted 11 lines · Jan Beulich
Quoted 6 lines · Bobby Eshleman

I followed with v2 feedback and attempted to convert the PE/COFF header into C instead of ASM. Unfortunately, this was only possible for the first part (Legacy) of the PE/COFF header. The other parts required addresses only available at link time (such as __2M_rwdata_end, __pe_SizeOfImage, efi_mb_start address, etc...), which effectively ruled out C.

I don't follow the conclusion drawn, so would you mind going into further detail?

No problem at all, bad explanation on my part. The core issue is actually about the legality of casting 64-bit addresses to 32-bit values in constant expressions, which then is sometimes complained about by GCC in terms of load-time computability...

Taking __2M_rwdata_end as an example. Attempting to use it in the PE/COFF optional header in C looks something like:

extern const char __2M_rwdata_end[];
extern const char efi_pe_head_end[];

struct optional_header optional_header = {
...

    .code_size = (uint32_t)((unsigned long)&__2M_rwdata_end) -
                    (uint32_t)((unsigned long)&efi_pe_head_end,

...
}

GCC throws "error: initializer element is not constant" because casting a 64-bit address to a 32-bit value is not a legal constant expression for static storage class objects, even though we know that in practice the address wouldn't ever be above 4GB.

efi_pe_head_end could potentially be calculated by header struct sizes, but doing that predictably yields the same error.

If we drop the explicit casting, GCC throws 'error: initializer element is not computable at load time'.

tl;dr:

I could not find a way to derive code size (data sections and all) without using a symbol location, which is an illegal constant expression operand for initializing static storage class objects... and I could not find a way to define the header in C without using an object of static storage class (global variable or static variable).

-Bob

#26 lore raw
Quoted 36 lines · Bobby Eshleman
Quoted 11 lines · Jan Beulich
Quoted 6 lines · Bobby Eshleman

I followed with v2 feedback and attempted to convert the PE/COFF header into C instead of ASM. Unfortunately, this was only possible for the first part (Legacy) of the PE/COFF header. The other parts required addresses only available at link time (such as __2M_rwdata_end, __pe_SizeOfImage, efi_mb_start address, etc...), which effectively ruled out C.

I don't follow the conclusion drawn, so would you mind going into further detail?

No problem at all, bad explanation on my part. The core issue is actually about the legality of casting 64-bit addresses to 32-bit values in constant expressions, which then is sometimes complained about by GCC in terms of load-time computability...

Taking __2M_rwdata_end as an example. Attempting to use it in the PE/COFF optional header in C looks something like:

extern const char __2M_rwdata_end[];
extern const char efi_pe_head_end[];

struct optional_header optional_header = {
...

    .code_size = (uint32_t)((unsigned long)&__2M_rwdata_end) -
                    (uint32_t)((unsigned long)&efi_pe_head_end,

...
}

GCC throws "error: initializer element is not constant" because casting a 64-bit address to a 32-bit value is not a legal constant expression for static storage class objects, even though we know that in practice the address wouldn't ever be above 4GB.

Quoted 5 lines

efi_pe_head_end could potentially be calculated by header struct sizes, but doing that predictably yields the same error.

If we drop the explicit casting, GCC throws 'error: initializer element is not computable at load time'.

Ah yes, I see now, and I'm aware of the compiler shortcoming. Even with the not really necessary uint32_t casts dropped the problem would still be there. So for your description this means it's not "required addresses only available at link time" but "required differences of addresses not computable or expressable at compile time".

Jan

Quoted 10 lines

tl;dr:

I could not find a way to derive code size (data sections and all) without using a symbol location, which is an illegal constant expression operand for initializing static storage class objects... and I could not find a way to define the header in C without using an object of static storage class (global variable or static variable).

-Bob

#27 lore raw
In reply to Bobby Eshleman

Hey all,

I just wanted to request more feedback on this series and put it on the radar, while acknowledging that I'm sure given the recent code freeze it is a busy time for everybody.

Best,
Bob

#28 lore raw
Quoted 2 lines · Bobby Eshleman

I just wanted to request more feedback on this series and put it on the radar, while acknowledging that I'm sure given the recent code freeze it is a busy time for everybody.

It is on my list of things to look at. While probably not a good excuse, my looking at previous versions of this makes we somewhat hesitant to open any of these patch mails ... But I mean to get to it.

Jan

#29 lore raw
Quoted 6 lines · Jan Beulich

It is on my list of things to look at. While probably not a good excuse, my looking at previous versions of this makes we somewhat hesitant to open any of these patch mails ... But I mean to get to it.

Jan

Thanks for this response. I did comb through your v2 feedback point-by-point and incorporate it into the code, so I do hope that ends up helping.

Best,
Bob