powerel7
/
base
2
0
Fork 0
Code Issues Pull Requests Releases Activity

gnu-efi package update 

Signed-off-by: basebuilder_pel7x64builder0 <basebuilder@powerel.org>
master
basebuilder_pel7x64builder0 2018-08-08 19:03:59 +02:00
parent 12118581b6
commit 44fb971b39
13 changed files with 878 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

33
SOURCES/0001-Mark-our-explicit-fall-through-so-Wextra-will-work-i.patch Normal file
View File

@ -0,0 +1,33 @@
From f3d3ef07eb69072b8bd2b0c5d4e6243ea38ecec9 Mon Sep 17 00:00:00 2001
From: Peter Jones <pjones@redhat.com>
Date: Thu, 2 Feb 2017 13:51:27 -0500
Subject: [PATCH 01/10] Mark our explicit fall through so -Wextra will work in
gcc 7

gcc 7 introduces detection of fall-through behavior in switch/case
statements, and will warn if -Wimplicit-fallthrough is present and there
is no comment stating that the fall-through is intentional. This is
also triggered by -Wextra, as it enables -Wimplicit-fallthrough=1.

This patch adds the comment in the one place we use fall-through.

Signed-off-by: Peter Jones <pjones@redhat.com>
---
lib/print.c | 1 +
1 file changed, 1 insertion(+)

diff --git a/lib/print.c b/lib/print.c
index b8a9d38..cb732f0 100644
--- a/lib/print.c
+++ b/lib/print.c
@@ -1131,6 +1131,7 @@ Returns:
case 'X':
Item.Width = Item.Long ? 16 : 8;
Item.Pad = '0';
+ /* falls through */
case 'x':
ValueToHex (
Item.Scratch,
--
2.9.3

37
SOURCES/0003-Fix-arm-build-paths-in-the-makefile.patch Normal file
View File

@ -0,0 +1,37 @@
From fa85eb1b6ac4ceff1672c7152b6f842c2f2ff728 Mon Sep 17 00:00:00 2001
From: Peter Jones <pjones@redhat.com>
Date: Thu, 2 Feb 2017 14:31:25 -0500
Subject: [PATCH 03/10] Fix arm build paths in the makefile

Previous work was apparently done with arm-linux-gnueabi-gcc as a
cross-builder, but our armv7 builders have native gcc with the target as
armv7hl-linux-gnueabi, so we need to munge the arch there to get our arm
path.

Signed-off-by: Peter Jones <pjones@redhat.com>
---
Make.defaults | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/Make.defaults b/Make.defaults
index 51bd7d8..e4d2ac1 100755
--- a/Make.defaults
+++ b/Make.defaults
@@ -62,12 +62,12 @@ OBJCOPY := $(prefix)$(CROSS_COMPILE)objcopy
# Host/target identification
OS := $(shell uname -s)
-HOSTARCH := $(shell $(HOSTCC) -dumpmachine | cut -f1 -d- | sed s,i[3456789]86,ia32,)
+HOSTARCH := $(shell $(HOSTCC) -dumpmachine | cut -f1 -d- | sed -e s,i[3456789]86,ia32, -e 's,armv7.*,arm,' )
ARCH := $(HOSTARCH)
# Get ARCH from the compiler if cross compiling
ifneq ($(CROSS_COMPILE),)
- override ARCH := $(shell $(CC) -dumpmachine | cut -f1 -d-| sed s,i[3456789]86,ia32,)
+ override ARCH := $(shell $(CC) -dumpmachine | cut -f1 -d-| sed -e s,i[3456789]86,ia32, -e 's,armv7.*,arm,' )
endif
# FreeBSD (and possibly others) reports amd64 instead of x86_64
--
2.9.3

31
SOURCES/0004-Work-around-Werror-maybe-uninitialized-not-being-ver.patch Normal file
View File

@ -0,0 +1,31 @@
From 6f5781e191cca6c20a75b82cc467c2256cbb5901 Mon Sep 17 00:00:00 2001
From: Peter Jones <pjones@redhat.com>
Date: Thu, 2 Feb 2017 15:23:55 -0500
Subject: [PATCH 04/10] Work around -Werror=maybe-uninitialized not being very
bright.

The compiler doesn't believe the loop always executes at least once,
even though the data in the first array entry doesn't satisfy the exit
condition. So just initialize the thing to shut it up.

Signed-off-by: Peter Jones <pjones@redhat.com>
---
apps/route80h.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/apps/route80h.c b/apps/route80h.c
index 723dd85..bf550a1 100644
--- a/apps/route80h.c
+++ b/apps/route80h.c
@@ -102,7 +102,7 @@ efi_main (EFI_HANDLE image_handle, EFI_SYSTEM_TABLE *systab)
InitializeLib(image_handle, systab);
EFI_PCI_IO *pciio = NULL;
lpcif_t lpcif;
- EFI_STATUS rc;
+ EFI_STATUS rc = EFI_SUCCESS;
struct {
uint16_t vendor;
uint16_t device;
--
2.9.3

39
SOURCES/0005-Fix-a-sign-error-in-the-debughook-example-app.patch Normal file
View File

@ -0,0 +1,39 @@
From 397b06ff74f7fd8fc71f5991bdfbfa6ef6da98fd Mon Sep 17 00:00:00 2001
From: Peter Jones <pjones@redhat.com>
Date: Thu, 2 Feb 2017 15:25:48 -0500
Subject: [PATCH 05/10] Fix a sign error in the debughook example app

On ISO C90 on i386 4294967294 is a signed integer, and so x can't be
greater (or equal) to that. Make it an unsigned and choose a better type
for the variable.

Signed-off-by: Peter Jones <pjones@redhat.com>
---
apps/debughook.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/apps/debughook.c b/apps/debughook.c
index fb6cdad..e1d1c50 100644
--- a/apps/debughook.c
+++ b/apps/debughook.c
@@ -45,7 +45,7 @@ DebugHook(void)
UINT8 *data = NULL;
UINTN dataSize = 0;
EFI_STATUS efi_status;
- register volatile UINTN x = 0;
+ register volatile unsigned long long x = 0;
extern char _text, _data;
if (x)
@@ -66,7 +66,7 @@ DebugHook(void)
while (x++) {
/* Make this so it can't /totally/ DoS us. */
#if defined(__x86_64__) || defined(__i386__) || defined(__i686__)
- if (x > 4294967294)
+ if (x > 4294967294ULL)
break;
__asm__ __volatile__("pause");
#elif defined(__aarch64__)
--
2.9.3

49
SOURCES/0006-Fix-typedef-of-EFI_PXE_BASE_CODE.patch Normal file
View File

@ -0,0 +1,49 @@
From e4ca94374c7ece8ec0100075710af8638e42c203 Mon Sep 17 00:00:00 2001
From: Peter Jones <pjones@redhat.com>
Date: Wed, 8 Feb 2017 15:28:18 -0500
Subject: [PATCH 06/10] Fix typedef of EFI_PXE_BASE_CODE

Commit 751cbce3 fixed up a bunch of types to better match the edk2
definitions and the names in the UEFI Spec, but while doing so
inadvertantly defined things thusly:

INTERFACE_DECL(_EFI_PXE_BASE_CODE_PROTOCOL);
...
typedef struct _EFI_PXE_BASE_CODE_PROTOCOL {
...
} EFI_PXE_BASE_CODE_PROTOCOL;
...
typedef struct _EFI_PXE_BASE_CODE_PROTOCOL _EFI_PXE_BASE_CODE;
typedef struct EFI_PXE_BASE_CODE_PROTOCOL EFI_PXE_BASE_CODE;

Because EFI_BASE_CODE_PROTOCOL is declared with a typedef, and is
therefore in the type namespace rather than the struct namespace, this
results in EFI_PXE_BASE_CODE being a forward declaration of an
incomplete type. The net result is that code which dereferences any
field in the struct, even with the correct names, will not correctly
build.

This patch changes both _EFI_PXE_BASE_CODE and EFI_PXE_BASE_CODE
typedefs to inherit from struct _EFI_PXE_BASE_CODE_PROTOCOL.

Signed-off-by: Peter Jones <pjones@redhat.com>
---
inc/efipxebc.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/inc/efipxebc.h b/inc/efipxebc.h
index 580a6ef..3760c7c 100644
--- a/inc/efipxebc.h
+++ b/inc/efipxebc.h
@@ -419,7 +419,7 @@ typedef struct _EFI_PXE_BASE_CODE_PROTOCOL {
// Use 'EFI_PXE_BASE_CODE_PROTOCOL_GUID' instead.
typedef struct _EFI_PXE_BASE_CODE_PROTOCOL _EFI_PXE_BASE_CODE;
-typedef struct EFI_PXE_BASE_CODE_PROTOCOL EFI_PXE_BASE_CODE;
+typedef struct _EFI_PXE_BASE_CODE_PROTOCOL EFI_PXE_BASE_CODE;
//
// Call Back Definitions
--
2.9.3

28
SOURCES/0007-make-clang-not-complain-about-fno-merge-constants.patch Normal file
View File

@ -0,0 +1,28 @@
From a01463c5035d2bacefc1ef7673b6ba2cc9815920 Mon Sep 17 00:00:00 2001
From: Peter Jones <pjones@redhat.com>
Date: Fri, 10 Feb 2017 15:16:42 -0500
Subject: [PATCH 07/10] make clang not complain about -fno-merge-constants

---
Make.defaults | 5 +++--
1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/Make.defaults b/Make.defaults
index e4d2ac1..cc52195 100755
--- a/Make.defaults
+++ b/Make.defaults
@@ -159,8 +159,9 @@ CFLAGS += $(ARCH3264) -g -O2 -Wall -Wextra -Werror \
else
CFLAGS += $(ARCH3264) -g -O2 -Wall -Wextra -Werror \
-fshort-wchar -fno-strict-aliasing \
- -fno-merge-constants -ffreestanding -fno-stack-protector \
- -fno-stack-check
+ -ffreestanding -fno-stack-protector \
+ -fno-stack-check \
+ $(if $(findstring gcc,$(CC)),-fno-merge-constants,)
endif
ARFLAGS += -U
--
2.9.3

26
SOURCES/0008-Fix-another-place-clang-complains-about.patch Normal file
View File

@ -0,0 +1,26 @@
From 0281df3c77fc4d8c67c0d23c656b4debed862989 Mon Sep 17 00:00:00 2001
From: Peter Jones <pjones@redhat.com>
Date: Fri, 10 Feb 2017 16:14:12 -0500
Subject: [PATCH 08/10] Fix another place clang complains about

Signed-off-by: Peter Jones <pjones@redhat.com>
---
lib/guid.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/lib/guid.c b/lib/guid.c
index bd1f1a5..6498e90 100644
--- a/lib/guid.c
+++ b/lib/guid.c
@@ -97,7 +97,7 @@ static struct {
{ &SMapId, L"ShellDevPathMap" },
{ &SAliasId, L"ShellAlias" },
- { NULL }
+ { NULL, L"" }
};
//
--
2.9.3

33
SOURCES/0009-route80h-remove-some-dead-code.patch Normal file
View File

@ -0,0 +1,33 @@
From ba05cb606c6c3a4ec9f60c588abef52e355e5c1a Mon Sep 17 00:00:00 2001
From: Peter Jones <pjones@redhat.com>
Date: Fri, 10 Feb 2017 16:14:55 -0500
Subject: [PATCH 09/10] route80h: remove some dead code

Signed-off-by: Peter Jones <pjones@redhat.com>
---
apps/route80h.c | 9 ---------
1 file changed, 9 deletions(-)

diff --git a/apps/route80h.c b/apps/route80h.c
index bf550a1..5272dd3 100644
--- a/apps/route80h.c
+++ b/apps/route80h.c
@@ -40,15 +40,6 @@ static inline void set_bit(volatile uint32_t *flag, int bit, int value)
Print(L"new value is 0x%2x\n", val);
}
-static inline int configspace_matches_ids(void *config, uint32_t vendor_id,
- uint32_t device_id)
-{
- uint32_t *cfg = config;
- if (cfg[0] == vendor_id && cfg[1] == device_id)
- return 1;
- return 0;
-}
-
static int is_device(EFI_PCI_IO *pciio, uint16_t vendor_id, uint16_t device_id)
{
lpcif_t lpcif;
--
2.9.3

57
SOURCES/0010-Make-clang-not-complain-about-the-debughook-s-optimi.patch Normal file
View File

@ -0,0 +1,57 @@
From a94906bd487c1a2e6d7827f31d7eb46bc09f3b43 Mon Sep 17 00:00:00 2001
From: Peter Jones <pjones@redhat.com>
Date: Fri, 10 Feb 2017 16:18:14 -0500
Subject: [PATCH 10/10] Make clang not complain about the debughook's
optimization settings

... still won't work, of course, because nobody at clang takes compiler
compatibility seriously while they keep implying that you can pretend
it's gcc.

Signed-off-by: Peter Jones <pjones@redhat.com>
---
Make.defaults | 3 ++-
apps/debughook.c | 5 +++--
2 files changed, 5 insertions(+), 3 deletions(-)

diff --git a/Make.defaults b/Make.defaults
index cc52195..a7778a7 100755
--- a/Make.defaults
+++ b/Make.defaults
@@ -161,7 +161,8 @@ CFLAGS += $(ARCH3264) -g -O2 -Wall -Wextra -Werror \
-fshort-wchar -fno-strict-aliasing \
-ffreestanding -fno-stack-protector \
-fno-stack-check \
- $(if $(findstring gcc,$(CC)),-fno-merge-constants,)
+ $(if $(findstring gcc,$(CC)),-fno-merge-constants,) \
+ $(if $(findstring clang,$(CC)),-Wno-unknown-pragmas,)
endif
ARFLAGS += -U
diff --git a/apps/debughook.c b/apps/debughook.c
index e1d1c50..fa36f62 100644
--- a/apps/debughook.c
+++ b/apps/debughook.c
@@ -37,8 +37,9 @@ GetVariable(CHAR16 *var, UINT8 **data, UINTN *len, EFI_GUID owner)
EFI_GUID DUMMY_GUID =
{0x55aad538, 0x8f82, 0x4e2a, {0xa4,0xf0,0xbe, 0x59, 0x13, 0xb6, 0x5f, 0x1e}};
+#pragma GCC push_options
+#pragma GCC optimize ("0")
static void
-__attribute__((__optimize__("0")))
DebugHook(void)
{
EFI_GUID guid = DUMMY_GUID;
@@ -81,7 +82,7 @@ DebugHook(void)
}
x = 1;
}
-
+#pragma GCC pop_options
EFI_STATUS
efi_main (EFI_HANDLE image, EFI_SYSTEM_TABLE *systab)
--
2.9.3

25
SOURCES/0011-Nerf-Werror-pragma-away.patch Normal file
View File

@ -0,0 +1,25 @@
From 6671a712db5c4dfecd4e7026057a99d9b1c4fe5e Mon Sep 17 00:00:00 2001
From: Peter Jones <pjones@redhat.com>
Date: Tue, 7 Mar 2017 10:11:19 -0500
Subject: [PATCH] Nerf -Werror=pragma away

---
Make.defaults | 2 ++
1 file changed, 2 insertions(+)

diff --git a/Make.defaults b/Make.defaults
index a7778a7..8882701 100755
--- a/Make.defaults
+++ b/Make.defaults
@@ -87,6 +87,8 @@ OBJDIR := $(TOPDIR)/$(ARCH)
# Arch-specific compilation flags
CPPFLAGS += -DCONFIG_$(ARCH)
+CFLAGS += -Wno-error=pragmas
+
ifeq ($(ARCH),ia64)
CFLAGS += -mfixed-range=f32-f127
endif
--
2.9.3

34
SOURCES/0012-Make-ia32-use-our-own-div-asm-on-gnu-C-as-well.patch Normal file
View File

@ -0,0 +1,34 @@
From bbd1ee0379327a58e042d288e371f314b7df2d55 Mon Sep 17 00:00:00 2001
From: Peter Jones <pjones@redhat.com>
Date: Wed, 15 Mar 2017 17:33:39 -0400
Subject: [PATCH] Make ia32 use our own div asm on gnu C as well.

---
lib/ia32/math.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/lib/ia32/math.c b/lib/ia32/math.c
index 81f51bf..fce7a8d 100644
--- a/lib/ia32/math.c
+++ b/lib/ia32/math.c
@@ -140,7 +140,7 @@ DivU64x32 (
// divide 64bit by 32bit and get a 64bit result
// N.B. only works for 31bit divisors!!
{
-#if defined(__GNUC__) && !defined(__MINGW32__)
+#if 0 && defined(__GNUC__) && !defined(__MINGW32__)
if (Remainder)
*Remainder = Dividend % Divisor;
return Dividend / Divisor;
@@ -157,7 +157,7 @@ DivU64x32 (
Rem = 0;
for (bit=0; bit < 64; bit++) {
-#ifdef __MINGW32__
+#if defined(__GNUC__) || defined(__MINGW32__)
asm (
"shll $1, %0\n\t"
"rcll $1, 4%0\n\t"
--
2.12.0

71
SOURCES/0013-Call-ar-in-deterministic-mode.patch Normal file
View File

@ -0,0 +1,71 @@
From 0e95c51225b01f0ca79dafd607ae6a3e404db6be Mon Sep 17 00:00:00 2001
From: Peter Jones <pjones@redhat.com>
Date: Tue, 28 Mar 2017 17:59:55 -0400
Subject: [PATCH 13/13] Call ar in deterministic mode.

We need the x86_64 and i686 builds of .a's to be the same, and that
means we need to not have timestamps. Also force the timestamps on disk
just in case that doesn't work, because RHEL's ar /silently ignores -D/.

v2: use "ar rvD" not "ar rv -D".

It's a wonder anybody ever gets these command line options right, if
"ar rv -D libfoo.a foo.o" doesn't use deterministic mode (or
complain), but "ar rvD libfoo.a foo.o" does.

v3: Add a bunch of junk to try to set timestamps to 0 manually

For some reason I'm still getting timestamps in the .a even though ar seems to
be invoked correctly. When I do "mock -r rhel-7-build --shell" and run make
manually, they're fine. Very strange.

v4: go back to v2, the problem isn't in the make process.

"ar rDv" works just fine, but /usr/lib/rpm/redhat/brp-strip-static-archive is
calling "%{__strip} -g $for_each.a", and it's rewriting our binary from
ts/uid/gid of 0/0/0 to $epoch/$UID/$GID. Awesomely /usr/bin/strip it seems to
have 3 modes of operation:
-U: the default, which adds $epoch/$UID/$GID to your binary archive
instead of just removing stuff. Clearly the Principle of Least
Surprise is strong here.
-p: preserve the timestamp from the original .a, but add UID and GID,
because this is 1980 and people use ar(1) for archiving stuff they
might want that out of.
-D: Condescend at you in a command line error and explain that -D both
is and is not a valid option:
/usr/bin/strip: invalid option -- 'D'
Usage: /usr/bin/strip <option(s)> in-file(s)
Removes symbols and sections from files
The options are:
...
-D --enable-deterministic-archives
Produce deterministic output when stripping archives
So I agree that it's invalid, but I think we may be pronouncing that
second vowel differently. They say in-VAL-id, I say IN-vuh-lid.

Nobody should ever have to run "strace -ttt -v -f -o make.strace make all",
just to discover the problem isn't even in there.

Related: rhbz#1310782

Signed-off-by: Peter Jones <pjones@redhat.com>
---
Make.defaults | 2 +-
1 files changed, 1 insertions(+), 1 deletion(-)

diff --git a/Make.defaults b/Make.defaults
index aa15c73..8f36365 100755
--- a/Make.defaults
+++ b/Make.defaults
@@ -167,7 +167,7 @@ CFLAGS += $(ARCH3264) -g -O2 -Wall -Wextra -Werror \
$(if $(findstring clang,$(CC)),-Wno-unknown-pragmas,)
endif
-ARFLAGS += -U
+ARFLAGS := rDv
ASFLAGS += $(ARCH3264)
LDFLAGS += -nostdlib --warn-common --no-undefined --fatal-warnings \
--build-id=sha1
--
2.9.3

415
SPECS/gnu-efi.spec Normal file
View File

@ -0,0 +1,415 @@
Summary: Development Libraries and headers for EFI
Name: gnu-efi
Version: 3.0.5
Release: 9%{?dist}%{?buildid}
Epoch: 1
Group: Development/System
License: BSD
URL: ftp://ftp.hpl.hp.com/pub/linux-ia64
BuildRoot: %{_tmppath}/%{name}-%{version}-%{release}-root-%(%{__id_u} -n)
ExclusiveArch: x86_64 aarch64
BuildRequires: git
%ifarch x86_64
#BuildRequires: glibc32
BuildRequires: glibc-devel(x86-32)
%endif
Source: http://superb-dca2.dl.sourceforge.net/project/gnu-efi/gnu-efi-%{version}.tar.bz2

Patch0001: 0001-Mark-our-explicit-fall-through-so-Wextra-will-work-i.patch
Patch0002: 0002-Fix-some-types-gcc-doesn-t-like.patch
Patch0003: 0003-Fix-arm-build-paths-in-the-makefile.patch
Patch0004: 0004-Work-around-Werror-maybe-uninitialized-not-being-ver.patch
Patch0005: 0005-Fix-a-sign-error-in-the-debughook-example-app.patch
Patch0006: 0006-Fix-typedef-of-EFI_PXE_BASE_CODE.patch
Patch0007: 0007-make-clang-not-complain-about-fno-merge-constants.patch
Patch0008: 0008-Fix-another-place-clang-complains-about.patch
Patch0009: 0009-route80h-remove-some-dead-code.patch
Patch0010: 0010-Make-clang-not-complain-about-the-debughook-s-optimi.patch
Patch0011: 0011-Nerf-Werror-pragma-away.patch
Patch0012: 0012-Make-ia32-use-our-own-div-asm-on-gnu-C-as-well.patch
Patch0013: 0013-Call-ar-in-deterministic-mode.patch

%define debug_package %{nil}

# brp-strip-static-archive will senselessly /add/ timestamps and uid/gid
# data to our .a and make them not multilib clean if we don't have this.
# Note that if we don't have the shell quotes there, -p becomes $2 on its
# invocation, and so it completely ignores it.
#
# Also note that if we try to use -D as we should (so it doesn't add
# uid/gid), strip(1) from binutils-2.25.1-22.base.el7.x86_64 throws a
# syntax error.
#
# True story.
#
%global __strip "%{__strip} -p"

# Figure out the right file path to use
%global efidir %(eval echo $(grep ^ID= /etc/os-release | sed -e 's/^ID=//' -e 's/rhel/redhat/'))

%ifarch x86_64
%global efiarch x86_64
%endif
%ifarch aarch64
%global efiarch aarch64
%endif
%ifarch %{ix86}
%global efiarch ia32
%endif

%description
This package contains development headers and libraries for developing
applications that run under EFI (Extensible Firmware Interface).

%package devel
Summary: Development Libraries and headers for EFI
Group: Development/System
Obsoletes: gnu-efi < 1:3.0.2-1
Requires: gnu-efi

%description devel
This package contains development headers and libraries for developing
applications that run under EFI (Extensible Firmware Interface).

%package utils
Summary: Utilities for EFI systems
Group: Applications/System

%description utils
This package contains utilties for debugging and developing EFI systems.

%prep
%setup -q -n gnu-efi-%{version}
git init
git config user.email "gnu-efi-owner@redhat.com"
git config user.name "RHEL Ninjas"
git config sendemail.to "gnu-efi-owner@fedoraproject.org"
git add .
git commit -a -q -m "%{version} baseline."
git am %{patches} </dev/null
git config --unset user.email
git config --unset user.name

%build
# Package cannot build with %{?_smp_mflags}.
make
make apps
%ifarch x86_64
setarch linux32 -B make ARCH=ia32 PREFIX=%{_prefix} LIBDIR=%{_prefix}/lib
setarch linux32 -B make ARCH=ia32 PREFIX=%{_prefix} LIBDIR=%{_prefix}/lib apps
%endif

%install
rm -rf %{buildroot}

mkdir -p %{buildroot}/%{_libdir}/gnuefi
mkdir -p %{buildroot}/boot/efi/EFI/%{efidir}/%{efiarch}
make PREFIX=%{_prefix} LIBDIR=%{_libdir} INSTALLROOT=%{buildroot} install
mv %{buildroot}/%{_libdir}/*.lds %{buildroot}/%{_libdir}/*.o %{buildroot}/%{_libdir}/gnuefi
mv %{efiarch}/apps/{route80h.efi,modelist.efi} %{buildroot}/boot/efi/EFI/%{efidir}/%{efiarch}/

%ifarch x86_64
mkdir -p %{buildroot}/%{_prefix}/lib/gnuefi
mkdir -p %{buildroot}/boot/efi/EFI/%{efidir}/ia32

setarch linux32 -B make PREFIX=%{_prefix} LIBDIR=%{_prefix}/lib INSTALLROOT=%{buildroot} ARCH=ia32 install
mv %{buildroot}/%{_prefix}/lib/*.{lds,o} %{buildroot}/%{_prefix}/lib/gnuefi/
mv ia32/apps/{route80h.efi,modelist.efi} %{buildroot}/boot/efi/EFI/%{efidir}/ia32/
%endif

%clean
rm -rf %{buildroot}

%files
%{_prefix}/lib*/*

%files devel
%defattr(-,root,root,-)
%doc README.* ChangeLog
%{_includedir}/efi

%files utils
%dir /boot/efi/EFI/%{efidir}/
%attr(0644,root,root) /boot/efi/EFI/%{efidir}/*/*.efi

%changelog
* Thu Mar 30 2017 Peter Jones <pjones@redhat.com> - 3.0.5-9
- Just don't build the .i686 package at all. After a scratch build, it's
clear that "strip -p" is not good enough, because our different builders
have non-matching UIDs for the build process, and -p adds uid/gid to the
archive. So there's no way to fix the multiarch conflict here without
either fixing that or fixing strip(1) with:
https://sourceware.org/git/gitweb.cgi?p=binutils-gdb.git;a=patch;h=7a093a78
We don't strictly need the .i686 package anyway, since we've moved to
making the dependent binaries all build the ia32 bits on x86_64 for
other reasons. Related: rhbz#1310782

* Thu Mar 30 2017 Peter Jones <pjones@redhat.com> - 3.0.5-9
- One more attempt at nerfing timestamps. It's surprising how broken this
can be.

"ar rDv" works just fine, but
/usr/lib/rpm/redhat/brp-strip-static-archive is calling "%{__strip} -g
$for_each.a", and it's rewriting our binary from ts/uid/gid of 0/0/0 to
$epoch/$UID/$GID. Awesomely /usr/bin/strip it seems to have 3 modes of
operation:
-U: the default, which adds $epoch/$UID/$GID to your binary archive
instead of just removing stuff. Clearly the Principle of Least
Surprise is strong here.
-p: preserve the timestamp from the original .a, but add UID and GID,
because this is 1980 and people use ar(1) for archiving stuff they
might want that out of.
-D: Condescend at you in a command line error and explain that -D both
is and is not a valid option:
/usr/bin/strip: invalid option -- 'D'
Usage: /usr/bin/strip <option(s)> in-file(s)
Removes symbols and sections from files
The options are:
...
-D --enable-deterministic-archives
Produce deterministic output when stripping archives
So I agree that it's invalid, but I think we may be pronouncing that
second vowel differently. They say in-VAL-id, I say IN-vuh-lid.

Nobody should ever have to run "strace -ttt -v -f -o make.strace make
all", just to discover the problem isn't even in there.
Related: rhbz#1310782

* Tue Mar 28 2017 Peter Jones <pjones@redhat.com> - 3.0.5-8
- Nerf the timestamps on our .o files while building, because RHEL's ar(1) is
horrible and silently ignores the 'D' option. It's fine, I probably didn't
put it there for any reason.
Related: rhbz#1310782

* Tue Mar 28 2017 Peter Jones <pjones@redhat.com> - 3.0.5-7
- Call ar(1) in deterministic mode so our .a's are multipath clean.
Related: rhbz#1310782

* Mon Mar 20 2017 Peter Jones <pjones@redhat.com> - 3.0.5-6
- Also build the ia32 bits in a separate 32-bit package for other consumers.
Related: rhbz#1310782

* Wed Mar 15 2017 Peter Jones <pjones@redhat.com> - 3.0.5-5
- Fix a codegin bug that makes it want libgcc_s (but not know it) on ia32.
Related: rhbz#1310782

* Mon Mar 13 2017 Peter Jones <pjones@redhat.com> - 3.0.5-4
- Package the ia32 bits somewhat better.
Related: rhbz#1310782

* Mon Mar 13 2017 Peter Jones <pjones@redhat.com> - 3.0.5-3
- Include ia32 bits in the x86_64 packages instead of making a separate
32-bit package
Resolves: rhbz#1310782

* Mon Mar 06 2017 Peter Jones <pjones@redhat.com> - 3.0.5-2
- Fix some bugs in the 3.0.5 release.
Related: rhbz#1310782

* Thu Feb 02 2017 Peter Jones <pjones@redhat.com> - 3.0.5-1
- Update to 3.0.5
- Re-enable ia32 builds for the most hilarious changelog series...
Resolves: rhbz#1310782

* Mon Jun 15 2015 Peter Jones <pjones@redhat.com> - 3.0.2-2
- Fix .spec mismerge from upstream that causes ia32 to build.
Related: rhbz#1190191
Related: rhbz#1115843
Related: rhbz#1100048

* Mon Jun 15 2015 Peter Jones <pjones@redhat.com> - 3.0.2-1
- Update to 3.0.2
Related: rhbz#1190191
Related: rhbz#1115843
Related: rhbz#1100048
- Fix base package requirement on subpackages

* Fri Aug 22 2014 Kyle McMartin <kyle@fedoraproject.org> - 3.0w-0.1
- New upstream version 3.0w
- Add pjones' build fixes patch from that other distro.
- Enable AArch64

* Fri Dec 27 2013 Daniel Mach <dmach@redhat.com> - 3.0u-2
- Mass rebuild 2013-12-27

* Fri Oct 25 2013 Peter Jones <pjones@redhat.com> - 3.0u-1
- Reflect that we're not supporting this on anything but x86_64.
Related: rhbz#1017861

* Tue Sep 24 2013 Peter Jones <pjones@redhat.com> - 3.0u-0.1
- Update to 3.0u
Related: rhbz#996863
- Split out subpackages so -devel can be multilib
- Fix path in apps subpackage to vary by distro.

* Thu Jul 25 2013 Peter Jones <pjones@redhat.com> - 3.0q-3
- Revert to 3.0q
Related: rhbz#978766

* Fri Jun 07 2013 Peter Jones <pjones@redhat.com> - 3.0t-0.1
- Update to 3.0t
- Don't allow use of mmx or sse registers.

* Thu May 16 2013 Peter Jones <pjones@redhat.com> - 3.0s-2
- Update to 3.0s
Related: rhbz#963359

* Thu Feb 14 2013 Fedora Release Engineering <rel-eng@lists.fedoraproject.org> - 3.0q-2
- Rebuilt for https://fedoraproject.org/wiki/Fedora_19_Mass_Rebuild

* Fri Jul 27 2012 Matthew Garrett <mjg@redhat.com> - 3.0q-1
- Update to current upstream
- License change - GPLv2+ to BSD

* Thu Jul 19 2012 Fedora Release Engineering <rel-eng@lists.fedoraproject.org> - 3.0e-18
- Rebuilt for https://fedoraproject.org/wiki/Fedora_18_Mass_Rebuild

* Wed Apr 25 2012 Peter Jones <pjones@redhat.com> - 3.0e-17
- Align .reloc section as well to make secureboot work (mfleming)

* Fri Jan 13 2012 Fedora Release Engineering <rel-eng@lists.fedoraproject.org> - 3.0e-16
- Rebuilt for https://fedoraproject.org/wiki/Fedora_17_Mass_Rebuild

* Thu Aug 11 2011 Peter Jones <pjones@redhat.com> - 3.0e-15
- Correctly pad the stack when doing uefi calls
Related: rhbz#677468
- Add ability to write UEFI callbacks and drivers
- Add test harness for ABI Calling Conventions

* Thu Jun 16 2011 Peter Jones <pjones@redhat.com> - 3.0e-14
- Handle uninitialized GOP driver gracefully.

* Wed Feb 09 2011 Fedora Release Engineering <rel-eng@lists.fedoraproject.org> - 3.0e-13
- Rebuilt for https://fedoraproject.org/wiki/Fedora_15_Mass_Rebuild

* Fri Sep 10 2010 Peter Jones <pjones@redhat.com> - 3.0e-12
- Add "modelist.efi" test utility in apps/

* Mon Jul 26 2010 Peter Jones <pjones@redhat.com> - 3.0e-11
- Add PciIo headers.

* Fri Jul 23 2010 Peter Jones <pjones@redhat.com> - 3.0e-10
- Add UEFI 2.x boot services.

* Tue Aug 11 2009 Peter Jones <pjones@redhat.com> - 3.0e-9
- Change ExclusiveArch to reflect arch changes in repos.

* Fri Jul 24 2009 Fedora Release Engineering <rel-eng@lists.fedoraproject.org> - 3.0e-8
- Rebuilt for https://fedoraproject.org/wiki/Fedora_12_Mass_Rebuild

* Fri Apr 03 2009 Peter Jones <pjones@redhat.com> - 3.0e-7
- Use nickc's workaround for #492183

* Tue Mar 31 2009 Peter Jones <pjones@redhat.com> - 3.0e-6.1
- Make a test package for nickc.

* Thu Mar 12 2009 Chris Lumens <clumens@redhat.com> 3.0e-6
- Add IA64 back into the list of build arches (#489544).

* Mon Mar 02 2009 Peter Jones <pjones@redhat.com> - 3.0e-5
- Switch to i586 from i386.

* Tue Feb 24 2009 Fedora Release Engineering <rel-eng@lists.fedoraproject.org> - 3.0e-4
- Rebuilt for https://fedoraproject.org/wiki/Fedora_11_Mass_Rebuild

* Fri Feb 13 2009 Peter Jones <pjones@redhat.com> - 3.0e-3
- Pad sections out in the provided linker scripts to make sure they all of
some content.

* Fri Oct 03 2008 Peter Jones <pjones@redhat.com> - 3.0e-2
- Fix install paths on x86_64.

* Thu Oct 02 2008 Peter Jones <pjones@redhat.com> - 3.0e-1
- Update to 3.0e
- Fix relocation bug in 3.0e

* Tue Jul 29 2008 Tom "spot" Callaway <tcallawa@redhat.com> - 3.0d-6
- fix license tag

* Mon Jul 28 2008 Peter Jones <pjones@redhat.com> - 3.0d-5
- Remove ia64 palproc code since its license isn't usable.
- Remove ia64 from ExclusiveArch since it can't build...

* Thu Mar 27 2008 Peter Jones <pjones@redhat.com> - 3.0d-4
- Fix uefi_call_wrapper(x, 10, ...) .
- Add efi_main wrappers and EFI_CALL() macro so drivers are possible.

* Mon Feb 18 2008 Fedora Release Engineering <rel-eng@fedoraproject.org> - 3.0d-3
- Autorebuild for GCC 4.3

* Fri Jan 11 2008 Peter Jones <pjones@redhat.com> - 3.0d-2
- Get rid of a bogus #ifdef .

* Wed Dec 19 2007 Peter Jones <pjones@redhat.com> - 3.0d-1
- Update to 3.0d

* Tue Jun 12 2007 Chris Lumens <clumens@redhat.com> - 3.0c-2
- Fixes for package review (#225846).

* Wed Jul 12 2006 Jesse Keating <jkeating@redhat.com> - 3.0c-1.1
- rebuild

* Thu Apr 27 2006 Chris Lumens <clumens@redhat.com> 3.0c-1
- Upgrade to gnu-efi-3.0c.
- Enable build on i386.

* Tue Feb 07 2006 Jesse Keating <jkeating@redhat.com> - 3.0a-7.2
- rebuilt for new gcc4.1 snapshot and glibc changes

* Fri Dec 09 2005 Jesse Keating <jkeating@redhat.com>
- rebuilt

* Thu Mar 3 2005 Jeremy Katz <katzj@redhat.com> - 3.0a-7
- rebuild with gcc 4

* Tue Sep 21 2004 Jeremy Katz <katzj@redhat.com> - 3.0a-6
- add fix from Jesse Barnes for newer binutils (#129197)

* Tue Jun 15 2004 Elliot Lee <sopwith@redhat.com>
- rebuilt

* Wed Apr 21 2004 Jeremy Katz <katzj@redhat.com> - 3.0a-4
- actually add the patch

* Tue Apr 20 2004 Bill Nottingham <notting@redhat.com> 3.0a-3
- add patch to coalesce some relocations (#120080, <erikj@sgi.com>)

* Fri Feb 13 2004 Elliot Lee <sopwith@redhat.com>
- rebuilt

* Fri Oct 4 2002 Jeremy Katz <katzj@redhat.com>
- rebuild in new environment

* Sun Jul 8 2001 Bill Nottingham <notting@redhat.com>
- update to 3.0

* Tue Jun 5 2001 Bill Nottingham <notting@redhat.com>
- add fix for invocations from the boot manager menu (#42222)

* Tue May 22 2001 Bill Nottingham <notting@redhat.com>
- add bugfix for efibootmgr (<schwab@suse.de>)

* Mon May 21 2001 Bill Nottingham <notting@redhat.com>
- update to 2.5
- add in efibootmgr from Dell (<Matt_Domsch@dell.com>)

* Thu May 3 2001 Bill Nottingham <notting@redhat.com>
- fix booting of kernels with extra arguments (#37711)

* Wed Apr 25 2001 Bill Nottingham <notting@redhat.com>
- take out Stephane's initrd patch

* Fri Apr 20 2001 Bill Nottingham <notting@redhat.com>
- fix the verbosity patch to not break passing arguments to images

* Wed Apr 18 2001 Bill Nottingham <notting@redhat.com>
- update to 2.0, build elilo, obsolete eli

* Tue Dec 5 2000 Bill Nottingham <notting@redhat.com>
- update to 1.1

* Thu Oct 26 2000 Bill Nottingham <notting@redhat.com>
- add patch for new toolchain, update to 1.0

* Thu Aug 17 2000 Bill Nottingham <notting@redhat.com>
- update to 0.9