[PATCH 0/6] Add AutoFDO and Propeller support for Clang build

[PATCH 0/6] Add AutoFDO and Propeller support for Clang build

[Rong Xu]

Hi,

This patch series is to integrate AutoFDO and Propeller support into
the Linux kernel. AutoFDO is a profile-guided optimization technique
that leverages hardware sampling to enhance binary performance.
Unlike Instrumentation-based FDO (iFDO), AutoFDO offers a user-friendly
and straightforward application process. While iFDO generally yields
superior profile quality and performance, our findings reveal that
AutoFDO achieves remarkable effectiveness, bringing performance close
to iFDO for benchmark applications. Similar to AutoFDO, Propeller too
utilizes hardware sampling to collect profiles and apply post-link
optimizations to improve the benchmark’s performance over and above
AutoFDO.

Our empirical data demonstrates significant performance improvements
with AutoFDO and Propeller, up to 10% on microbenchmarks and up to 5%
on large warehouse-scale benchmarks. This makes a strong case for their
inclusion as supported features in the upstream kernel.

Background

A significant fraction of fleet processing cycles (excluding idle time)
from data center workloads are attributable to the kernel. Ware-house
scale workloads maximize performance by optimizing the production kernel
using iFDO (a.k.a instrumented PGO, Profile Guided Optimization).

iFDO can significantly enhance application performance but its use
within the kernel has raised concerns. AutoFDO is a variant of FDO that
uses the hardware’s Performance Monitoring Unit (PMU) to collect
profiling data. While AutoFDO typically yields smaller performance
gains than iFDO, it presents unique benefits for optimizing kernels.

AutoFDO eliminates the need for instrumented kernels, allowing a single
optimized kernel to serve both execution and profile collection. It also
minimizes slowdown during profile collection, potentially yielding
higher-fidelity profiling, especially for time-sensitive code, compared
to iFDO. Additionally, AutoFDO profiles can be obtained from production
environments via the hardware’s PMU whereas iFDO profiles require
carefully curated load tests that are representative of real-world
traffic.

AutoFDO facilitates profile collection across diverse targets.
Preliminary studies indicate significant variation in kernel hot spots
within Google’s infrastructure, suggesting potential performance gains
through target-specific kernel customization.

Furthermore, other advanced compiler optimization techniques, including
ThinLTO and Propeller can be stacked on top of AutoFDO, similar to iFDO.
ThinLTO achieves better runtime performance through whole-program
analysis and cross module optimizations. The main difference between
traditional LTO and ThinLTO is that the latter is scalable in time and
memory. Propeller is a profile-guided, post-link optimizer that improves
the performance of large-scale applications compiled with LLVM. It
operates by relinking the binary based on an additional round of runtime
profiles, enabling precise optimizations that are not possible at
compile time.

This patch series adds AutoFDO and Propeller support to the kernel. The
actual solution comes in six parts:

[P 1] Add the build support for using AutoFDO in Clang

      Add the basic support for AutoFDO build and provide the
      instructions for using AutoFDO.

[P 2] Fix objtool for bogus warnings when -ffunction-sections is enabled

[P 3] Change the subsection ordering when -ffunction-sections is enabled

[P 4] Enable –ffunction-sections for the AutoFDO build

[P 5] Enable Machine Function Split (MFS) optimization for AutoFDO

[P 6] Add Propeller configuration to the kernel build

Patch 1 provides basic AutoFDO build support. Patches 2 to 5 further
enhance the performance of AutoFDO builds and are functionally dependent
on Patch 1. Patch 6 enables support for Propeller and is dependent on
patch 2 and patch 3.

Caveats

AutoFDO is compatible with both GCC and Clang, but the patches in this
series are exclusively applicable to LLVM 17 or newer for AutoFDO and
LLVM 19 or newer for Propeller. For profile conversion, two different
tools could be used, llvm_profgen or create_llvm_prof. llvm_profgen
needs to be the LLVM 19 or newer, or just the LLVM trunk. Alternatively,
create_llvm_prof v0.30.1 or newer can be used instead of llvm-profgen.

Additionally, the build is only supported on x86 platforms equipped
with PMU capabilities, such as LBR on Intel machines. More
specifically:
 * Intel platforms: works on every platform that supports LBR;
   we have tested on Skylake.
 * AMD platforms: tested on AMD Zen3 with the BRS feature. The kernel
   needs to be configured with “CONFIG_PERF_EVENTS_AMD_BRS=y", To
   check, use
   $ cat /proc/cpuinfo | grep “ brs”
   For the AMD Zen4, AMD LBRV2 is supported, but we suspect a bug with
   AMD LBRv2 implementation in Genoa which blocks the usage.

Experiments and Results

Experiments were conducted to compare the performance of AutoFDO-optimized
kernel images (version 6.9.x) against default builds.. The evaluation
encompassed both open source microbenchmarks and real-world production
services from Google and Meta. The selected microbenchmarks included Neper,
a network subsystem benchmark, and UnixBench which is a comprehensive suite
for assessing various kernel operations.

For Neper, AutoFDO optimization resulted in a 6.1% increase in throughput
and a 10.6% reduction in latency. Unixbench saw a 2.2% improvement in its
index score under low system load and a 2.6% improvement under high system
load.

For further details on the improvements observed in Google and Meta's
production services, please refer to the LLVM discourse post:
https://discourse.llvm.org/t/optimizing-the-linux-kernel-with-autofdo-including-thinlto-and-propeller/79108

Thanks,

Rong Xu and Han Shen

Rong Xu (6):
  Add AutoFDO support for Clang build
  objtool: Fix unreachable instruction warnings for weak funcitons
  Change the symbols order when --ffuntion-sections is enabled
  AutoFDO: Enable -ffunction-sections for the AutoFDO build
  AutoFDO: Enable machine function split optimization for AutoFDO
  Add Propeller configuration for kernel build.

 Documentation/dev-tools/autofdo.rst   | 203 ++++++++++++++++++++++++++
 Documentation/dev-tools/index.rst     |   2 +
 Documentation/dev-tools/propeller.rst | 188 ++++++++++++++++++++++++
 MAINTAINERS                           |  14 ++
 Makefile                              |   2 +
 arch/Kconfig                          |  42 ++++++
 arch/x86/Kconfig                      |   2 +
 arch/x86/boot/compressed/Makefile     |   3 +
 arch/x86/kernel/vmlinux.lds.S         |   4 +
 arch/x86/platform/efi/Makefile        |   1 +
 drivers/firmware/efi/libstub/Makefile |   2 +
 include/asm-generic/vmlinux.lds.h     |  56 ++++++-
 scripts/Makefile.autofdo              |  22 +++
 scripts/Makefile.lib                  |  20 +++
 scripts/Makefile.propeller            |  25 ++++
 tools/objtool/check.c                 |   2 +
 tools/objtool/elf.c                   |  13 +-
 17 files changed, 588 insertions(+), 13 deletions(-)
 create mode 100644 Documentation/dev-tools/autofdo.rst
 create mode 100644 Documentation/dev-tools/propeller.rst
 create mode 100644 scripts/Makefile.autofdo
 create mode 100644 scripts/Makefile.propeller


base-commit: 9903efbddba0d14133b5a3c75088b558d2e34ac3
-- 
2.46.0.rc1.232.g9752f9e123-goog

[PATCH 1/6] Add AutoFDO support for Clang build

[Rong Xu]

Add the build support for using Clang's AutoFDO. Building the kernel
with AutoFDO does not reduce the optimization level from the
compiler. AutoFDO uses hardware sampling to gather information about
the frequency of execution of different code paths within a binary.
This information is then used to guide the compiler's optimization
decisions, resulting in a more efficient binary. Experiments
showed that the kernel can improve up to 10% in latency.

The support requires a Clang compiler after LLVM 17. This submission
is limited to x86 platforms that support PMU features like LBR on
Intel machines and AMD Zen3 BRS. Support for SPE on ARM 1,
 and BRBE on ARM 1 is part of planned future work.

Here is an example workflow for AutoFDO kernel:

1) Build the kernel on the HOST machine with LLVM enabled, for example,
       $ make menuconfig LLVM=1
    Turn on AutoFDO build config:
      CONFIG_AUTOFDO_CLANG=y
    With a configuration that has LLVM enabled, use the following
    command:
       scripts/config -e AUTOFDO_CLANG
    After getting the config, build with
      $ make LLVM=1

2) Install the kernel on the TEST machine.

3) Run the load tests. The '-c' option in perf specifies the sample
   event period. We suggest     using a suitable prime number,
   like 500009, for this purpose.
   For Intel platforms:
      $ perf record -e BR_INST_RETIRED.NEAR_TAKEN:k -a -N -b -c <count> \
        -o <perf_file> -- <loadtest>
   For AMD platforms:
      The supported system are: Zen3 with BRS, or Zen4 with amd_lbr_v2
     For Zen3:
      $ cat proc/cpuinfo | grep " brs"
      For Zen4:
      $ cat proc/cpuinfo | grep amd_lbr_v2
      $ perf record --pfm-events RETIRED_TAKEN_BRANCH_INSTRUCTIONS:k -a \
        -N -b -c <count> -o <perf_file> -- <loadtest>

4) (Optional) Download the raw perf file to the HOST machine.

5) To generate an AutoFDO profile, two offline tools are available:
   create_llvm_prof and llvm_profgen. The create_llvm_prof tool is part
   of the AutoFDO project and can be found on GitHub
   (https://github.com/google/autofdo), version v0.30.1 or later. The
   llvm_profgen tool is included in the LLVM compiler itself. It's
   important to note that the version of llvm_profgen doesn't need to
   match the version of Clang. It needs to be the LLVM 19 release or
   later, or from the LLVM trunk.
      $ llvm-profgen --kernel --binary=<vmlinux> --perfdata=<perf_file> \
        -o <profile_file>
   or
      $ create_llvm_prof --binary=<vmlinux> --profile=<perf_file> \
        --format=extbinary -o <profile_file>

   Note that multiple AutoFDO profile files can be merged into one via:
      $ llvm-profdata merge -o <profile_file>  <profile_1> ... <profile_n>

6) Rebuild the kernel using the AutoFDO profile file with the same config
   as step 1, (Note CONFIG_AUTOFDO_CLANG needs to be enabled):
      $ make LLVM=1 CLANG_AUTOFDO_PROFILE=<profile_file>

Co-developed-by: Han Shen <shenhan@google.com>
Signed-off-by: Han Shen <shenhan@google.com>
Signed-off-by: Rong Xu <xur@google.com>
Suggested-by: Sriraman Tallam <tmsriram@google.com>
Suggested-by: Krzysztof Pszeniczny <kpszeniczny@google.com>
Suggested-by: Nick Desaulniers <ndesaulniers@google.com>
Suggested-by: Stephane Eranian <eranian@google.com>
---
 Documentation/dev-tools/autofdo.rst | 203 ++++++++++++++++++++++++++++
 Documentation/dev-tools/index.rst   |   1 +
 MAINTAINERS                         |   7 +
 Makefile                            |   1 +
 arch/Kconfig                        |  20 +++
 arch/x86/Kconfig                    |   1 +
 scripts/Makefile.autofdo            |  20 +++
 scripts/Makefile.lib                |  10 ++
 tools/objtool/check.c               |   1 +
 9 files changed, 264 insertions(+)
 create mode 100644 Documentation/dev-tools/autofdo.rst
 create mode 100644 scripts/Makefile.autofdo

diff --git a/Documentation/dev-tools/autofdo.rst b/Documentation/dev-tools/autofdo.rst
new file mode 100644
index 000000000000..e3f7ee8ee6bb
--- /dev/null
+++ b/Documentation/dev-tools/autofdo.rst
@@ -0,0 +1,203 @@
+.. SPDX-License-Identifier: GPL-2.0
+
+===================================
+Using AutoFDO with the Linux kernel
+===================================
+
+This enables AutoFDO build support for the kernel when using
+the Clang compiler. AutoFDO (Auto-Feedback-Directed Optimization)
+is a type of profile-guided optimization (PGO) used to enhance the
+performance of binary executables. It gathers information about the
+frequency of execution of various code paths within a binary using
+hardware sampling. This data is then used to guide the compiler's
+optimization decisions, resulting in a more efficient binary. AutoFDO
+is a powerful optimization technique, and data indicates that it can
+significantly improve kernel performance. It's especially beneficial
+for workloads affected by front-end stalls.
+
+For AutoFDO builds, unlike non-FDO builds, the user must supply a
+profile. Acquiring an AutoFDO profile can be done in several ways.
+AutoFDO profiles are created by converting hardware sampling using
+the "perf" tool. It is crucial that the workload used to create these
+perf files is representative; they must exhibit runtime
+characteristics similar to the workloads that are intended to be
+optimized. Failure to do so will result in the compiler optimizing
+for the wrong objective.
+
+The AutoFDO profile often encapsulates the program's behavior. If the
+performance-critical codes are architecture-independent, the profile
+can be applied across platforms to achieve performance gains. For
+instance, using the profile generated on Intel architecture to build
+a kernel for AMD architecture can also yield performance improvements.
+
+There are two methods for acquiring a representative profile:
+(1) Sample real workloads using a production environment.
+(2) Generate the profile using a representative load test.
+When enabling the AutoFDO build configuration without providing an
+AutoFDO profile, the compiler only modifies the dwarf information in
+the kernel without impacting runtime performance. It's advisable to
+use a kernel binary built with the same AutoFDO configuration to
+collect the perf profile. While it's possible to use a kernel built
+with different options, it may result in inferior performance.
+
+One can collect profiles using AutoFDO build for the previous kernel.
+AutoFDO employs relative line numbers to match the profiles, offering
+some tolerance for source changes. This mode is commonly used in a
+production environment for profile collection.
+
+In a profile collection based on a load test, the AutoFDO collection
+process consists of the following steps:
+
+#. Initial build: The kernel is built with AutoFDO options
+   without a profile.
+
+#. Profiling: The above kernel is then run with a representative
+   workload to gather execution frequency data. This data is
+   collected using hardware sampling, via perf. AutoFDO is most
+   effective on platforms supporting advanced PMU features like
+   LBR on Intel machines.
+
+#. AutoFDO profile generation: Perf output file is converted to
+   the AutoFDO profile via offline tools.
+
+The support requires a Clang compiler LLVM 17 or later.
+
+Preparation
+===========
+
+Configure the kernel with:
+
+   .. code-block:: make
+
+      CONFIG_AUTOFDO_CLANG=y
+
+Customization
+=============
+
+You can enable or disable AutoFDO build for individual file and directories by
+adding a line similar to the following to the respective kernel Makefile:
+
+- For enabling a single file (e.g. foo.o)
+
+     .. code-block:: make
+
+        AUTOFDO_PROFILE_foo.o := y
+
+- For enabling all files in one directory
+
+     .. code-block:: make
+
+        AUTOFDO_PROFILE := y
+
+- For disabling one file
+
+     .. code-block:: make
+
+        AUTOFDO_PROFILE_foo.o := n
+
+- For disabling all files in one directory
+
+     .. code-block:: make
+
+        AUTOFDO_PROFILE := n
+
+
+Workflow
+========
+
+Here is an example workflow for AutoFDO kernel:
+
+
+
+1)  Build the kernel on the HOST machine with LLVM enabled, for example,
+
+      .. code-block:: make
+
+         $ make menuconfig LLVM=1
+
+
+    Turn on AutoFDO build config:
+
+      .. code-block:: make
+
+         CONFIG_AUTOFDO_CLANG=y
+
+    With a configuration that with LLVM enabled, use the following command:
+
+      .. code-block:: sh
+
+         $ scripts/config -e AUTOFDO_CLANG
+
+    After getting the config, build with
+
+      .. code-block:: make
+
+         $ make LLVM=1
+
+2) Install the kernel on the TEST machine.
+
+3) Run the load tests. The '-c' option in perf specifies the sample
+   event period. We suggest using a suitable prime number, like 500009,
+   for this purpose.
+
+   - For Intel platforms:
+
+      .. code-block:: sh
+
+         $ perf record -e BR_INST_RETIRED.NEAR_TAKEN:k -a -N -b -c <count> -o <perf_file> -- <loadtest>
+
+   - For AMD platforms: For Intel platforms:
+     The supported systems are: Zen3 with BRS, or Zen4 with amd_lbr_v2. To check,
+     For Zen3:
+
+      .. code-block:: sh
+
+         $ cat proc/cpuinfo | grep " brs"
+
+      For Zen4:
+
+      .. code-block:: sh
+
+         $ cat proc/cpuinfo | grep amd_lbr_v2
+
+      The following command generated the perf data file:
+
+      .. code-block:: sh
+
+         $ perf record --pfm-events RETIRED_TAKEN_BRANCH_INSTRUCTIONS:k -a -N -b \
+           -c <count> -o <perf_file> -- <loadtest>
+
+4) (Optional) Download the raw perf file to the HOST machine.
+
+5) To generate an AutoFDO profile, two offline tools are available:
+   create_llvm_prof and llvm_profgen. The create_llvm_prof tool is part
+   of the AutoFDO project and can be found on GitHub
+   (https://github.com/google/autofdo),  version v0.30.1 or later.
+   The llvm_profgen tool is included in the LLVM compiler itself. It's
+   important to note that the version of llvm_profgen doesn't need to match
+   the version of Clang. It needs to be the LLVM 19 release of Clang
+   or later, or just from the LLVM trunk.
+
+      .. code-block:: sh
+
+         $ llvm-profgen --kernel --binary=<vmlinux> --perfdata=<perf_file> -o <profile_file>
+
+   or
+      .. code-block:: sh
+
+         $ create_llvm_prof --binary=<vmlinux> --profile=<perf_file> --format=extbinary -o <profile_file>
+
+   Note that multiple AutoFDO profile files can be merged into one via:
+
+      .. code-block:: sh
+
+         $ llvm-profdata merge -o <profile_file>  <profile_1> <profile_2> ... <profile_n>
+
+
+6) Rebuild the kernel using the AutoFDO profile file with the same config as step 1,
+    (Note CONFIG_AUTOFDO_CLANG needs to be enabled):
+
+      .. code-block:: sh
+
+         $ make LLVM=1 CLANG_AUTOFDO_PROFILE=<profile_file
+
diff --git a/Documentation/dev-tools/index.rst b/Documentation/dev-tools/index.rst
index efa49cdc8e2e..46636e4efe15 100644
--- a/Documentation/dev-tools/index.rst
+++ b/Documentation/dev-tools/index.rst
@@ -32,6 +32,7 @@ Documentation/dev-tools/testing-overview.rst
    kunit/index
    ktap
    checkuapi
+   autofdo
 
 
 .. only::  subproject and html
diff --git a/MAINTAINERS b/MAINTAINERS
index cd2ca0c3158e..8a89e7f0d9d5 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -3500,6 +3500,13 @@ F:	kernel/audit*
 F:	lib/*audit.c
 K:	\baudit_[a-z_0-9]\+\b
 
+AUTOFDO BUILD
+M:	Rong Xu <xur@google.com>
+M:	Han Shen <shenhan@google.com>
+S:	Supported
+F:	Documentation/dev-tools/autofdo.rst
+F:	scripts/Makefile.autofdo
+
 AUXILIARY BUS DRIVER
 M:	Greg Kroah-Hartman <gregkh@linuxfoundation.org>
 R:	Dave Ertman <david.m.ertman@intel.com>
diff --git a/Makefile b/Makefile
index 06aa6402b385..5ae30cc94a26 100644
--- a/Makefile
+++ b/Makefile
@@ -1024,6 +1024,7 @@ include-$(CONFIG_UBSAN)		+= scripts/Makefile.ubsan
 include-$(CONFIG_KCOV)		+= scripts/Makefile.kcov
 include-$(CONFIG_RANDSTRUCT)	+= scripts/Makefile.randstruct
 include-$(CONFIG_GCC_PLUGINS)	+= scripts/Makefile.gcc-plugins
+include-$(CONFIG_AUTOFDO_CLANG)	+= scripts/Makefile.autofdo
 
 include $(addprefix $(srctree)/, $(include-y))
 
diff --git a/arch/Kconfig b/arch/Kconfig
index 975dd22a2dbd..e12599c4ab63 100644
--- a/arch/Kconfig
+++ b/arch/Kconfig
@@ -802,6 +802,26 @@ config LTO_CLANG_THIN
 	  If unsure, say Y.
 endchoice
 
+config ARCH_SUPPORTS_AUTOFDO_CLANG
+	bool
+
+config AUTOFDO_CLANG
+	bool "Enable Clang's AutoFDO build (EXPERIMENTAL)"
+	depends on ARCH_SUPPORTS_AUTOFDO_CLANG
+	depends on CC_IS_CLANG && CLANG_VERSION >= 170000
+	help
+	  This option enables Clang’s AutoFDO build. When
+	  an AutoFDO profile is specified in variable
+	  CLANG_AUTOFDO_PROFILE during the build process,
+	  Clang uses the profile to optimize the kernel.
+
+	  If no profile is specified, AutoFDO options are
+	  still passed to Clang to facilitate the collection
+	  of perf data for creating an AutoFDO profile in
+	  subsequent builds.
+
+	  If unsure, say N.
+
 config ARCH_SUPPORTS_CFI_CLANG
 	bool
 	help
diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig
index 1d7122a1883e..dca526b1364f 100644
--- a/arch/x86/Kconfig
+++ b/arch/x86/Kconfig
@@ -122,6 +122,7 @@ config X86
 	select ARCH_USES_CFI_TRAPS		if X86_64 && CFI_CLANG
 	select ARCH_SUPPORTS_LTO_CLANG
 	select ARCH_SUPPORTS_LTO_CLANG_THIN
+	select ARCH_SUPPORTS_AUTOFDO_CLANG
 	select ARCH_USE_BUILTIN_BSWAP
 	select ARCH_USE_CMPXCHG_LOCKREF		if X86_CMPXCHG64
 	select ARCH_USE_MEMTEST
diff --git a/scripts/Makefile.autofdo b/scripts/Makefile.autofdo
new file mode 100644
index 000000000000..882fa7e801ba
--- /dev/null
+++ b/scripts/Makefile.autofdo
@@ -0,0 +1,20 @@
+# SPDX-License-Identifier: GPL-2.0
+
+# Enable available and selected Clang AutoFDO features.
+
+CFLAGS_AUTOFDO_CLANG := -fdebug-info-for-profiling -mllvm -enable-fs-discriminator=true -mllvm -improved-fs-discriminator=true
+
+ifdef CLANG_AUTOFDO_PROFILE
+CFLAGS_AUTOFDO_CLANG += -fprofile-sample-use=$(CLANG_AUTOFDO_PROFILE)
+endif
+
+ifdef CONFIG_LTO_CLANG
+ifdef CONFIG_LTO_CLANG_THIN
+ifdef CLANG_AUTOFDO_PROFILE
+KBUILD_LDFLAGS += --lto-sample-profile=$(CLANG_AUTOFDO_PROFILE)
+endif
+KBUILD_LDFLAGS += --mllvm=-enable-fs-discriminator=true --mllvm=-improved-fs-discriminator=true -plugin-opt=thinlto
+endif
+endif
+
+export CFLAGS_AUTOFDO_CLANG
diff --git a/scripts/Makefile.lib b/scripts/Makefile.lib
index 9f06f6aaf7fc..c2cab5adaf25 100644
--- a/scripts/Makefile.lib
+++ b/scripts/Makefile.lib
@@ -209,6 +209,16 @@ _c_flags += $(if $(patsubst n%,, \
 	-D__KCSAN_INSTRUMENT_BARRIERS__)
 endif
 
+#
+# Enable Clang's AutoFDO build flags for a file or directory depending on
+# variables AUTOFDO_PROFILE_obj.o and AUTOFDO_PROFILE.
+#
+ifeq ($(CONFIG_AUTOFDO_CLANG),y)
+_c_flags += $(if $(patsubst n%,, \
+	$(AUTOFDO_PROFILE_$(target-stem).o)$(AUTOFDO_PROFILE)y), \
+	$(CFLAGS_AUTOFDO_CLANG))
+endif
+
 # $(src) for including checkin headers from generated source files
 # $(obj) for including generated headers from checkin source files
 ifeq ($(KBUILD_EXTMOD),)
diff --git a/tools/objtool/check.c b/tools/objtool/check.c
index 0a33d9195b7a..254913498c3c 100644
--- a/tools/objtool/check.c
+++ b/tools/objtool/check.c
@@ -4488,6 +4488,7 @@ static int validate_ibt(struct objtool_file *file)
 		    !strcmp(sec->name, "__jump_table")			||
 		    !strcmp(sec->name, "__mcount_loc")			||
 		    !strcmp(sec->name, ".kcfi_traps")			||
+		    !strcmp(sec->name, ".llvm.call-graph-profile")	||
 		    strstr(sec->name, "__patchable_function_entries"))
 			continue;
 
-- 
2.46.0.rc1.232.g9752f9e123-goog

[PATCH 2/6] objtool: Fix unreachable instruction warnings for weak funcitons

[Rong Xu]

In the presence of both weak and strong function definitions, the
linker drops the weak symbol in favor of a strong symbol, but
leaves the code in place. Code in ignore_unreachable_insn() has
some heuristics to suppress the warning, but it does not work when
-ffunction-sections is enabled.

Suppose function foo has both strong and weak definitions.
Case 1: The strong definition has an annotated section name,
like .init.text. Only the weak definition will be placed into
.text.foo. But since the section has no symbols, there will be no
"hole" in the section.

Case 2: Both sections are without an annotated section name.
Both will be placed into .text.foo section, but there will be only one
symbol (the strong one). If the weak code is before the strong code,
there is no "hole" as it fails to find the right-most symbol before
the offset.

The fix is to use the first node to compute the hole if hole.sym
is empty. If there is no symbol in the section, the first node
will be NULL, in which case, -1 is returned to skip the whole
section.

Co-developed-by: Han Shen <shenhan@google.com>
Signed-off-by: Han Shen <shenhan@google.com>
Signed-off-by: Rong Xu <xur@google.com>
Suggested-by: Sriraman Tallam <tmsriram@google.com>
Suggested-by: Krzysztof Pszeniczny <kpszeniczny@google.com>
---
 tools/objtool/elf.c | 13 ++++++++-----
 1 file changed, 8 insertions(+), 5 deletions(-)

diff --git a/tools/objtool/elf.c b/tools/objtool/elf.c
index 3d27983dc908..fa88bb254ccc 100644
--- a/tools/objtool/elf.c
+++ b/tools/objtool/elf.c
@@ -224,12 +224,15 @@ int find_symbol_hole_containing(const struct section *sec, unsigned long offset)
 	if (n)
 		return 0; /* not a hole */
 
-	/* didn't find a symbol for which @offset is after it */
-	if (!hole.sym)
-		return 0; /* not a hole */
+	/*
+	 * @offset >= sym->offset + sym->len, find symbol after it.
+	 * Use the first node in rb_tree when hole.sym is NULL.
+	 */
+	if (hole.sym)
+		n = rb_next(&hole.sym->node);
+	else
+		n = rb_first_cached(&sec->symbol_tree);
 
-	/* @offset >= sym->offset + sym->len, find symbol after it */
-	n = rb_next(&hole.sym->node);
 	if (!n)
 		return -1; /* until end of address space */
 
-- 
2.46.0.rc1.232.g9752f9e123-goog

[PATCH 3/6] Change the symbols order when --ffuntion-sections is enabled

[Rong Xu]

When the -ffunction-sections compiler option is enabled, each function
is placed in a separate section named .text.function_name rather than
putting all functions in a single .text section.

However, using -function-sections can cause problems with the
linker script. The comments included in include/asm-generic/vmlinux.lds.h
note these issues.:
  “TEXT_MAIN here will match .text.fixup and .text.unlikely if dead
   code elimination is enabled, so these sections should be converted
   to use ".." first.”

It is unclear whether there is a straightforward method for converting
a suffix to "..". This patch modifies the order of subsections within the
text output section when the -ffunction-sections flag is enabled.
Specifically, it repositions sections with certain fixed patterns (for
example .text.unlikely) before TEXT_MAIN, ensuring that they are grouped
and matched together.

Note that the limitation arises because the linker script employs glob
patterns instead of regular expressions for string matching. While there
is a method to maintain the current order using complex patterns, this
significantly complicates the pattern and increases the likelihood of
errors.

Co-developed-by: Han Shen <shenhan@google.com>
Signed-off-by: Han Shen <shenhan@google.com>
Signed-off-by: Rong Xu <xur@google.com>
Suggested-by: Sriraman Tallam <tmsriram@google.com>
Suggested-by: Krzysztof Pszeniczny <kpszeniczny@google.com>
---
 include/asm-generic/vmlinux.lds.h | 19 ++++++++++++++++---
 1 file changed, 16 insertions(+), 3 deletions(-)

diff --git a/include/asm-generic/vmlinux.lds.h b/include/asm-generic/vmlinux.lds.h
index 5703526d6ebf..f3de66bda293 100644
--- a/include/asm-generic/vmlinux.lds.h
+++ b/include/asm-generic/vmlinux.lds.h
@@ -582,9 +582,21 @@
  * during second ld run in second ld pass when generating System.map
  *
  * TEXT_MAIN here will match .text.fixup and .text.unlikely if dead
- * code elimination is enabled, so these sections should be converted
- * to use ".." first.
+ * code elimination or function-section is enabled. Match these symbols
+ * first when in these builds.
  */
+#if defined(CONFIG_LD_DEAD_CODE_DATA_ELIMINATION) || defined(CONFIG_LTO_CLANG)
+#define TEXT_TEXT							\
+		*(.text.asan.* .text.tsan.*)				\
+		*(.text.unknown .text.unknown.*)			\
+		*(.text.unlikely .text.unlikely.*)			\
+		ALIGN_FUNCTION();					\
+		*(.text.hot .text.hot.*)				\
+		*(TEXT_MAIN .text.fixup)				\
+		NOINSTR_TEXT						\
+		*(.ref.text)						\
+	MEM_KEEP(init.text*)
+#else
 #define TEXT_TEXT							\
 		ALIGN_FUNCTION();					\
 		*(.text.hot .text.hot.*)				\
@@ -594,7 +606,8 @@
 		NOINSTR_TEXT						\
 		*(.ref.text)						\
 		*(.text.asan.* .text.tsan.*)				\
-	MEM_KEEP(init.text*)						\
+	MEM_KEEP(init.text*)
+#endif
 
 
 /* sched.text is aling to function alignment to secure we have same
-- 
2.46.0.rc1.232.g9752f9e123-goog

[PATCH 4/6] AutoFDO: Enable -ffunction-sections for the AutoFDO build

[Rong Xu]

Enable -ffunction-sections by default for the AutoFDO build.

With -ffunction-sections, the compiler places each function in its own
section named .text.function_name instead of placing all functions in
the .text section. In the AutoFDO build, this allows the linker to
utilize profile information to reorganize functions for improved
utilization of iCache and iTLB.

Co-developed-by: Han Shen <shenhan@google.com>
Signed-off-by: Han Shen <shenhan@google.com>
Signed-off-by: Rong Xu <xur@google.com>
Suggested-by: Sriraman Tallam <tmsriram@google.com>
---
 include/asm-generic/vmlinux.lds.h | 37 ++++++++++++++++++++++++-------
 scripts/Makefile.autofdo          |  2 +-
 2 files changed, 30 insertions(+), 9 deletions(-)

diff --git a/include/asm-generic/vmlinux.lds.h b/include/asm-generic/vmlinux.lds.h
index f3de66bda293..97c8399e5532 100644
--- a/include/asm-generic/vmlinux.lds.h
+++ b/include/asm-generic/vmlinux.lds.h
@@ -95,18 +95,25 @@
  * With LTO_CLANG, the linker also splits sections by default, so we need
  * these macros to combine the sections during the final link.
  *
+ * With LTO_CLANG, the linker also splits sections by default, so we need
+ * these macros to combine the sections during the final link.
+ *
  * RODATA_MAIN is not used because existing code already defines .rodata.x
  * sections to be brought in with rodata.
  */
-#if defined(CONFIG_LD_DEAD_CODE_DATA_ELIMINATION) || defined(CONFIG_LTO_CLANG)
+#if defined(CONFIG_LD_DEAD_CODE_DATA_ELIMINATION) || defined(CONFIG_LTO_CLANG) || \
+defined(CONFIG_AUTOFDO_CLANG)
 #define TEXT_MAIN .text .text.[0-9a-zA-Z_]*
+#else
+#define TEXT_MAIN .text
+#endif
+#if defined(CONFIG_LD_DEAD_CODE_DATA_ELIMINATION) || defined(CONFIG_LTO_CLANG)
 #define DATA_MAIN .data .data.[0-9a-zA-Z_]* .data..L* .data..compoundliteral* .data.$__unnamed_* .data.$L*
 #define SDATA_MAIN .sdata .sdata.[0-9a-zA-Z_]*
 #define RODATA_MAIN .rodata .rodata.[0-9a-zA-Z_]* .rodata..L*
 #define BSS_MAIN .bss .bss.[0-9a-zA-Z_]* .bss..compoundliteral*
 #define SBSS_MAIN .sbss .sbss.[0-9a-zA-Z_]*
 #else
-#define TEXT_MAIN .text
 #define DATA_MAIN .data
 #define SDATA_MAIN .sdata
 #define RODATA_MAIN .rodata
@@ -577,6 +584,20 @@
 		__cpuidle_text_end = .;					\
 		__noinstr_text_end = .;
 
+#ifdef CONFIG_AUTOFDO_CLANG
+#define TEXT_HOT							\
+		__hot_text_start = .;					\
+		*(.text.hot .text.hot.*)				\
+		__hot_text_end = .;
+#define TEXT_UNLIKELY							\
+		__unlikely_text_start = .;				\
+		*(.text.unlikely .text.unlikely.*)			\
+		__unlikely_text_end = .;
+#else
+#define TEXT_HOT *(.text.hot .text.hot.*)
+#define TEXT_UNLIKELY *(.text.unlikely .text.unlikely.*)
+#endif
+
 /*
  * .text section. Map to function alignment to avoid address changes
  * during second ld run in second ld pass when generating System.map
@@ -585,13 +606,14 @@
  * code elimination or function-section is enabled. Match these symbols
  * first when in these builds.
  */
-#if defined(CONFIG_LD_DEAD_CODE_DATA_ELIMINATION) || defined(CONFIG_LTO_CLANG)
+#if defined(CONFIG_LD_DEAD_CODE_DATA_ELIMINATION) || defined(CONFIG_LTO_CLANG) || \
+defined(CONFIG_AUTOFDO_CLANG)
 #define TEXT_TEXT							\
 		*(.text.asan.* .text.tsan.*)				\
 		*(.text.unknown .text.unknown.*)			\
-		*(.text.unlikely .text.unlikely.*)			\
+		TEXT_UNLIKELY						\
 		ALIGN_FUNCTION();					\
-		*(.text.hot .text.hot.*)				\
+		TEXT_HOT						\
 		*(TEXT_MAIN .text.fixup)				\
 		NOINSTR_TEXT						\
 		*(.ref.text)						\
@@ -599,9 +621,9 @@
 #else
 #define TEXT_TEXT							\
 		ALIGN_FUNCTION();					\
-		*(.text.hot .text.hot.*)				\
+		TEXT_HOT						\
 		*(TEXT_MAIN .text.fixup)				\
-		*(.text.unlikely .text.unlikely.*)			\
+		TEXT_UNLIKELY						\
 		*(.text.unknown .text.unknown.*)			\
 		NOINSTR_TEXT						\
 		*(.ref.text)						\
@@ -609,7 +631,6 @@
 	MEM_KEEP(init.text*)
 #endif
 
-
 /* sched.text is aling to function alignment to secure we have same
  * address even at second ld pass when generating System.map */
 #define SCHED_TEXT							\
diff --git a/scripts/Makefile.autofdo b/scripts/Makefile.autofdo
index 882fa7e801ba..f765bd9e81d7 100644
--- a/scripts/Makefile.autofdo
+++ b/scripts/Makefile.autofdo
@@ -5,7 +5,7 @@
 CFLAGS_AUTOFDO_CLANG := -fdebug-info-for-profiling -mllvm -enable-fs-discriminator=true -mllvm -improved-fs-discriminator=true
 
 ifdef CLANG_AUTOFDO_PROFILE
-CFLAGS_AUTOFDO_CLANG += -fprofile-sample-use=$(CLANG_AUTOFDO_PROFILE)
+CFLAGS_AUTOFDO_CLANG += -fprofile-sample-use=$(CLANG_AUTOFDO_PROFILE) -ffunction-sections
 endif
 
 ifdef CONFIG_LTO_CLANG
-- 
2.46.0.rc1.232.g9752f9e123-goog

[PATCH 5/6] AutoFDO: Enable machine function split optimization for AutoFDO

[Rong Xu]

Enable the machine function split optimization for AutoFDO in Clang.

Machine function split (MFS) is a pass in the Clang compiler that
splits a function into hot and cold parts. The linker groups all
cold blocks across functions together. This decreases hot code
fragmentation and improves iCache and iTLB utilization.

MFS requires a profile so this is enabled only for the AutoFDO builds.

Co-developed-by: Han Shen <shenhan@google.com>
Signed-off-by: Han Shen <shenhan@google.com>
Signed-off-by: Rong Xu <xur@google.com>
Suggested-by: Sriraman Tallam <tmsriram@google.com>
Suggested-by: Krzysztof Pszeniczny <kpszeniczny@google.com>
---
 include/asm-generic/vmlinux.lds.h | 6 ++++++
 scripts/Makefile.autofdo          | 2 ++
 2 files changed, 8 insertions(+)

diff --git a/include/asm-generic/vmlinux.lds.h b/include/asm-generic/vmlinux.lds.h
index 97c8399e5532..7d9dc8a3c046 100644
--- a/include/asm-generic/vmlinux.lds.h
+++ b/include/asm-generic/vmlinux.lds.h
@@ -593,9 +593,14 @@ defined(CONFIG_AUTOFDO_CLANG)
 		__unlikely_text_start = .;				\
 		*(.text.unlikely .text.unlikely.*)			\
 		__unlikely_text_end = .;
+#define TEXT_SPLIT							\
+		__split_text_start = .;					\
+		*(.text.split .text.split.[0-9a-zA-Z_]*)		\
+		__split_text_end = .;
 #else
 #define TEXT_HOT *(.text.hot .text.hot.*)
 #define TEXT_UNLIKELY *(.text.unlikely .text.unlikely.*)
+#define TEXT_SPLIT
 #endif
 
 /*
@@ -611,6 +616,7 @@ defined(CONFIG_AUTOFDO_CLANG)
 #define TEXT_TEXT							\
 		*(.text.asan.* .text.tsan.*)				\
 		*(.text.unknown .text.unknown.*)			\
+		TEXT_SPLIT						\
 		TEXT_UNLIKELY						\
 		ALIGN_FUNCTION();					\
 		TEXT_HOT						\
diff --git a/scripts/Makefile.autofdo b/scripts/Makefile.autofdo
index f765bd9e81d7..80ad06689947 100644
--- a/scripts/Makefile.autofdo
+++ b/scripts/Makefile.autofdo
@@ -6,6 +6,7 @@ CFLAGS_AUTOFDO_CLANG := -fdebug-info-for-profiling -mllvm -enable-fs-discriminat
 
 ifdef CLANG_AUTOFDO_PROFILE
 CFLAGS_AUTOFDO_CLANG += -fprofile-sample-use=$(CLANG_AUTOFDO_PROFILE) -ffunction-sections
+CFLAGS_AUTOFDO_CLANG += -fsplit-machine-functions
 endif
 
 ifdef CONFIG_LTO_CLANG
@@ -14,6 +15,7 @@ ifdef CLANG_AUTOFDO_PROFILE
 KBUILD_LDFLAGS += --lto-sample-profile=$(CLANG_AUTOFDO_PROFILE)
 endif
 KBUILD_LDFLAGS += --mllvm=-enable-fs-discriminator=true --mllvm=-improved-fs-discriminator=true -plugin-opt=thinlto
+KBUILD_LDFLAGS += -plugin-opt=-split-machine-functions
 endif
 endif
 
-- 
2.46.0.rc1.232.g9752f9e123-goog

[PATCH 6/6] Add Propeller configuration for kernel build.

[Rong Xu]

Add the build support for using Clang's Propeller optimizer. Like
AutoFDO, Propeller uses hardware sampling to gather information
about the frequency of execution of different code paths within a
binary. This information is then used to guide the compiler's
optimization decisions, resulting in a more efficient binary.

The support requires a Clang compiler LLVM 19 or later, and the
create_llvm_prof tool
(https://github.com/google/autofdo/releases/tag/v0.30.1). This
submission is limited to x86 platforms that support PMU features
like LBR on Intel machines and AMD Zen3 BRS.

For Arm, we plan to send patches for SPE-based Propeller when
AutoFDO for Arm is ready.

Here is an example workflow for building an AutoFDO+Propeller
optimized kernel:

1) Build the kernel on the HOST machine, with AutoFDO and Propeller
   build config
      CONFIG_AUTOFDO_CLANG=y
      CONFIG_PROPELLER_CLANG=y
   then
      $ make LLVM=1 CLANG_AUTOFDO_PROFILE=<autofdo_profile>

“<autofdo_profile>” is the profile collected when doing a non-Propeller
AutoFDO build. This step builds a kernel that has the same optimization
level as AutoFDO, plus a metadata section that records basic block
information. This kernel image runs as fast as an AutoFDO optimized
kernel.

2) Install the kernel on test/production machines.

3) Run the load tests. The '-c' option in perf specifies the sample
   event period. We suggest using a suitable prime number,
   like 500009, for this purpose.
   For Intel platforms:
      $ perf record -e BR_INST_RETIRED.NEAR_TAKEN:k -a -N -b -c <count> \
        -o <perf_file> -- <loadtest>
   For AMD platforms:
      The supported system are: Zen3 with BRS, or Zen4 with amd_lbr_v2
      # To see if Zen3 support LBR:
      $ cat proc/cpuinfo | grep " brs"
      # To see if Zen4 support LBR:
      $ cat proc/cpuinfo | grep amd_lbr_v2
      # If the result is yes, then collect the profile using:
      $ perf record --pfm-events RETIRED_TAKEN_BRANCH_INSTRUCTIONS:k -a \
        -N -b -c <count> -o <perf_file> -- <loadtest>

4) (Optional) Download the raw perf file to the HOST machine.

5) Generate Propeller profile:
   $ create_llvm_prof --binary=<vmlinux> --profile=<perf_file> \
     --format=propeller --propeller_output_module_name \
     --out=<propeller_profile_prefix>_cc_profile.txt \
     --propeller_symorder=<propeller_profile_prefix>_ld_profile.txt

   “create_llvm_prof” is the profile conversion tool, and a prebuilt
   binary for linux can be found on
   https://github.com/google/autofdo/releases/tag/v0.30.1 (can also build
   from source).

   "<propeller_profile_prefix>" can be something like
   "/home/user/dir/any_string".

   This command generates a pair of Propeller profiles:
   "<propeller_profile_prefix>_cc_profile.txt" and
   "<propeller_profile_prefix>_ld_profile.txt".

6) Rebuild the kernel using the AutoFDO and Propeller profile files.
      CONFIG_AUTOFDO_CLANG=y
      CONFIG_PROPELLER_CLANG=y
   and
      $ make LLVM=1 CLANG_AUTOFDO_PROFILE=<autofdo_profile> \
        CLANG_PROPELLER_PROFILE_PREFIX=<propeller_profile_prefix>

Co-developed-by: Han Shen <shenhan@google.com>
Signed-off-by: Han Shen <shenhan@google.com>
Signed-off-by: Rong Xu <xur@google.com>
Suggested-by: Sriraman Tallam <tmsriram@google.com>
Suggested-by: Krzysztof Pszeniczny <kpszeniczny@google.com>
Suggested-by: Nick Desaulniers <ndesaulniers@google.com>
Suggested-by: Stephane Eranian <eranian@google.com>
---
 Documentation/dev-tools/index.rst     |   1 +
 Documentation/dev-tools/propeller.rst | 188 ++++++++++++++++++++++++++
 MAINTAINERS                           |   7 +
 Makefile                              |   1 +
 arch/Kconfig                          |  22 +++
 arch/x86/Kconfig                      |   1 +
 arch/x86/boot/compressed/Makefile     |   3 +
 arch/x86/kernel/vmlinux.lds.S         |   4 +
 arch/x86/platform/efi/Makefile        |   1 +
 drivers/firmware/efi/libstub/Makefile |   2 +
 include/asm-generic/vmlinux.lds.h     |   8 +-
 scripts/Makefile.lib                  |  10 ++
 scripts/Makefile.propeller            |  25 ++++
 tools/objtool/check.c                 |   1 +
 14 files changed, 270 insertions(+), 4 deletions(-)
 create mode 100644 Documentation/dev-tools/propeller.rst
 create mode 100644 scripts/Makefile.propeller

diff --git a/Documentation/dev-tools/index.rst b/Documentation/dev-tools/index.rst
index 46636e4efe15..16e33eadb73b 100644
--- a/Documentation/dev-tools/index.rst
+++ b/Documentation/dev-tools/index.rst
@@ -33,6 +33,7 @@ Documentation/dev-tools/testing-overview.rst
    ktap
    checkuapi
    autofdo
+   propeller
 
 
 .. only::  subproject and html
diff --git a/Documentation/dev-tools/propeller.rst b/Documentation/dev-tools/propeller.rst
new file mode 100644
index 000000000000..15ef0e6d973e
--- /dev/null
+++ b/Documentation/dev-tools/propeller.rst
@@ -0,0 +1,188 @@
+.. SPDX-License-Identifier: GPL-2.0
+
+=====================================
+Using Propeller with the Linux kernel
+=====================================
+
+This enables Propeller build support for the kernel when using Clang
+compiler. Propeller is a profile-guided optimization (PGO) method used
+to optimize binary executables. Like AutoFDO, it utilizes hardware
+sampling to gather information about the frequency of execution of
+different code paths within a binary. Unlike AutoFDO, this information
+is then used right before linking phase to optimize (among others)
+block layout within and across functions.
+
+A few important notes about adopting Propeller optimization:
+
+#. Although it can be used as a standalone optimization step, it is
+   strongly recommended to apply Propeller on top of AutoFDO,
+   AutoFDO+ThinLTO or Instrument FDO. The rest of this document
+   assumes this paradigm.
+
+#. Propeller uses another round of profiling on top of
+   AutoFDO/AutoFDO+ThinLTO/iFDO. The whole build process involves
+   "build-afdo - train-afdo - build-propeller - train-propeller -
+   build-optimized".
+
+#. Propeller requires LLVM 19 release or later for Clang/Clang++
+   and the linker(ld.lld).
+
+#. In addition to LLVM toolchain, Propeller requires a profiling
+   conversion tool: https://github.com/google/autofdo with a release
+   after v0.30.1: https://github.com/google/autofdo/releases/tag/v0.30.1.
+
+The Propeller optimization process involves the following steps:
+
+#. Initial building: Build the AutoFDO or AutoFDO+ThinLTO binary as
+   you would normally do, but with a set of compile-time / link-time
+   flags, so that a special metadata section is created within the
+   kernel binary. The special section is only intend to be used by the
+   profiling tool, it is not part of the runtime image, nor does it
+   change kernel run time text sections.
+
+#. Profiling: The above kernel is then run with a representative
+   workload to gather execution frequency data. This data is collected
+   using hardware sampling, via perf. Propeller is most effective on
+   platforms supporting advanced PMU features like LBR on Intel
+   machines. This step is the same as profiling the kernel for AutoFDO
+   (the exact perf parameters can be different).
+
+#. Propeller profile generation: Perf output file is converted to a
+   pair of Propeller profiles via an offline tool.
+
+#. Optimized build: Build the AutoFDO or AutoFDO+ThinLTO optimized
+   binary as you would normally do, but with a compile-time /
+   link-time flag to pick up the Propeller compile time and link time
+   profiles. This build step uses 3 profiles - the AutoFDO profile,
+   the Propeller compile-time profile and the Propeller link-time
+   profile.
+
+#. Deployment: The optimized kernel binary is deployed and used
+   in production environments, providing improved performance
+   and reduced latency.
+
+Preparation
+===========
+
+Configure the kernel with:
+
+   .. code-block:: make
+
+      CONFIG_AUTOFDO_CLANG=y
+      CONFIG_PROPELLER_CLANG=y
+
+Customization
+=============
+
+You can enable or disable Propeller build for individual file and
+directories by adding a line similar to the following to the
+respective kernel Makefile:
+
+- For enabling a single file (e.g. foo.o)
+
+     .. code-block:: make
+
+        PROPELLER_PROFILE_foo.o := y
+
+- For enabling all files in one directory
+
+     .. code-block:: make
+
+        PROPELLER_PROFILE := y
+
+- For disabling one file
+
+     .. code-block:: make
+
+        PROPELLER_PROFILE_foo.o := n
+
+- For disabling all files in one directory
+
+     .. code-block:: make
+
+        PROPELLER__PROFILE := n
+
+
+Workflow
+========
+
+Here is an example workflow for building an AutoFDO+Propeller kernel:
+
+1) Assuming an AutoFDO profile is already collected following
+   instructions in the AutoFDO document, build the kernel on the HOST
+   machine, with AutoFDO and Propeller build configs:
+
+      .. code-block:: make
+
+         CONFIG_AUTOFDO_CLANG=y
+         CONFIG_PROPELLER_CLANG=y
+
+   and
+
+      .. code-block:: sh
+
+         $ make LLVM=1 CLANG_AUTOFDO_PROFILE=<autofdo-profile-name>
+
+2) Install the kernel on the TEST machine.
+
+3) Run the load tests. The '-c' option in perf specifies the sample
+   event period. We suggest using a suitable prime number, like 500009,
+   for this purpose.
+
+   - For Intel platforms:
+
+      .. code-block:: sh
+
+         $ perf record -e BR_INST_RETIRED.NEAR_TAKEN:k -a -N -b -c \
+           <count> -o <perf_file> -- <loadtest>
+
+   - For AMD platforms:
+
+      .. code-block:: sh
+
+         $ perf record --pfm-event RETIRED_TAKEN_BRANCH_INSTRUCTIONS:k \
+           -a -N -b -c <count> -o <perf_file> -- <loadtest>
+
+   Note you can repeat the above steps to collect multiple <perf_file>s.
+
+4) (Optional) Download the raw perf file(s) to the HOST machine.
+
+5) Use the create_llvm_prof tool (https://github.com/google/autofdo) to Generate Propeller profile.
+
+      .. code-block:: sh
+
+         $ create_llvm_prof --binary=<vmlinux> --profile=<perf_file> \
+                            --format=propeller --propeller_output_module_name \
+                            --out=<propeller_profile_prefix>_cc_profile.txt \
+                            --propeller_symorder=<propeller_profile_prefix>_ld_profile.txt
+
+   "<propeller_profile_prefix>" can be something like
+   "/home/user/dir/any_string".
+
+   This command generates a pair of Propeller profiles:
+   "<propeller_profile_prefix>_cc_profile.txt" and
+   "<propeller_profile_prefix>_ld_profile.txt".
+
+   If there are more than 1 perf_file collected in the previous step,
+   you can create a temp list file "<perf_file_list>" with each line
+   containing one perf file name and run:
+
+      .. code-block:: sh
+
+         $ create_llvm_prof --binary=<vmlinux> --profile=@<perf_file_list> \
+                            --format=propeller --propeller_output_module_name \
+                            --out=<propeller_profile_prefix>_cc_profile.txt \
+                            --propeller_symorder=<propeller_profile_prefix>_ld_profile.txt
+
+6) Rebuild the kernel using the AutoFDO and Propeller profiles.
+
+      .. code-block:: make
+
+         CONFIG_AUTOFDO_CLANG=y
+         CONFIG_PROPELLER_CLANG=y
+
+   and
+
+      .. code-block:: sh
+
+         $ make LLVM=1 CLANG_AUTOFDO_PROFILE=<profile_file> CLANG_PROPELLER_PROFILE_PREFIX=<propeller_profile_prefix>
diff --git a/MAINTAINERS b/MAINTAINERS
index 8a89e7f0d9d5..0c7f3cebe4fe 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -17974,6 +17974,13 @@ S:	Maintained
 F:	include/linux/psi*
 F:	kernel/sched/psi.c
 
+PROPELLER BUILD
+M:	Rong Xu <xur@google.com>
+M:	Han Shen <shenhan@google.com>
+S:	Supported
+F:	Documentation/dev-tools/propeller.rst
+F:	scripts/Makefile.propeller
+
 PRINTK
 M:	Petr Mladek <pmladek@suse.com>
 R:	Steven Rostedt <rostedt@goodmis.org>
diff --git a/Makefile b/Makefile
index 5ae30cc94a26..85a96d973f20 100644
--- a/Makefile
+++ b/Makefile
@@ -1025,6 +1025,7 @@ include-$(CONFIG_KCOV)		+= scripts/Makefile.kcov
 include-$(CONFIG_RANDSTRUCT)	+= scripts/Makefile.randstruct
 include-$(CONFIG_GCC_PLUGINS)	+= scripts/Makefile.gcc-plugins
 include-$(CONFIG_AUTOFDO_CLANG)	+= scripts/Makefile.autofdo
+include-$(CONFIG_PROPELLER_CLANG)	+= scripts/Makefile.propeller
 
 include $(addprefix $(srctree)/, $(include-y))
 
diff --git a/arch/Kconfig b/arch/Kconfig
index e12599c4ab63..5b136e904400 100644
--- a/arch/Kconfig
+++ b/arch/Kconfig
@@ -822,6 +822,28 @@ config AUTOFDO_CLANG
 
 	  If unsure, say N.
 
+config ARCH_SUPPORTS_PROPELLER_CLANG
+	bool
+
+config PROPELLER_CLANG
+	bool "Enable Clang's Propeller build"
+	depends on ARCH_SUPPORTS_PROPELLER_CLANG
+	depends on AUTOFDO_CLANG
+	depends on CC_IS_CLANG && CLANG_VERSION >= 190000
+	help
+	  This option enables Clang’s Propeller build which
+	  is on top of AutoFDO build. When the Propeller profiles
+	  is specified in variable CLANG_PROPELLER_PROFILE_PREFIX
+	  during the build process, Clang uses the profiles to
+	  optimize the kernel.
+
+	  If no profile is specified, Proepller options are
+	  still passed to Clang to facilitate the collection
+	  of perf data for creating the Propeller profiles in
+	  subsequent builds.
+
+	  If unsure, say N.
+
 config ARCH_SUPPORTS_CFI_CLANG
 	bool
 	help
diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig
index dca526b1364f..6fb5269d39b0 100644
--- a/arch/x86/Kconfig
+++ b/arch/x86/Kconfig
@@ -123,6 +123,7 @@ config X86
 	select ARCH_SUPPORTS_LTO_CLANG
 	select ARCH_SUPPORTS_LTO_CLANG_THIN
 	select ARCH_SUPPORTS_AUTOFDO_CLANG
+	select ARCH_SUPPORTS_PROPELLER_CLANG    if X86_64
 	select ARCH_USE_BUILTIN_BSWAP
 	select ARCH_USE_CMPXCHG_LOCKREF		if X86_CMPXCHG64
 	select ARCH_USE_MEMTEST
diff --git a/arch/x86/boot/compressed/Makefile b/arch/x86/boot/compressed/Makefile
index f2051644de94..35d19b4e6361 100644
--- a/arch/x86/boot/compressed/Makefile
+++ b/arch/x86/boot/compressed/Makefile
@@ -17,6 +17,9 @@
 #	(see scripts/Makefile.lib size_append)
 #	compressed vmlinux.bin.all + u32 size of vmlinux.bin.all
 
+# Do not run Propeller optimizer for early boot code.
+PROPELLER_PROFILE              := n
+
 targets := vmlinux vmlinux.bin vmlinux.bin.gz vmlinux.bin.bz2 vmlinux.bin.lzma \
 	vmlinux.bin.xz vmlinux.bin.lzo vmlinux.bin.lz4 vmlinux.bin.zst
 
diff --git a/arch/x86/kernel/vmlinux.lds.S b/arch/x86/kernel/vmlinux.lds.S
index 3509afc6a672..167dd05323cf 100644
--- a/arch/x86/kernel/vmlinux.lds.S
+++ b/arch/x86/kernel/vmlinux.lds.S
@@ -440,6 +440,10 @@ SECTIONS
 
 	STABS_DEBUG
 	DWARF_DEBUG
+#ifdef CONFIG_PROPELLER_CLANG
+	.llvm_bb_addr_map : { *(.llvm_bb_addr_map) }
+#endif
+
 	ELF_DETAILS
 
 	DISCARDS
diff --git a/arch/x86/platform/efi/Makefile b/arch/x86/platform/efi/Makefile
index 543df9a1379d..e0c846b6d636 100644
--- a/arch/x86/platform/efi/Makefile
+++ b/arch/x86/platform/efi/Makefile
@@ -1,6 +1,7 @@
 # SPDX-License-Identifier: GPL-2.0
 KASAN_SANITIZE := n
 GCOV_PROFILE := n
+PROPELLER_PROFILE := n
 
 obj-$(CONFIG_EFI) 		+= memmap.o quirks.o efi.o efi_$(BITS).o \
 				   efi_stub_$(BITS).o
diff --git a/drivers/firmware/efi/libstub/Makefile b/drivers/firmware/efi/libstub/Makefile
index 06f0428a723c..55ca5250df1a 100644
--- a/drivers/firmware/efi/libstub/Makefile
+++ b/drivers/firmware/efi/libstub/Makefile
@@ -56,6 +56,8 @@ KBUILD_CFLAGS := $(filter-out $(CC_FLAGS_CFI), $(KBUILD_CFLAGS))
 # disable LTO
 KBUILD_CFLAGS := $(filter-out $(CC_FLAGS_LTO), $(KBUILD_CFLAGS))
 
+PROPELLER_PROFILE		:= n
+
 lib-y				:= efi-stub-helper.o gop.o secureboot.o tpm.o \
 				   file.o mem.o random.o randomalloc.o pci.o \
 				   skip_spaces.o lib-cmdline.o lib-ctype.o \
diff --git a/include/asm-generic/vmlinux.lds.h b/include/asm-generic/vmlinux.lds.h
index 7d9dc8a3c046..ea3d8bf51edd 100644
--- a/include/asm-generic/vmlinux.lds.h
+++ b/include/asm-generic/vmlinux.lds.h
@@ -95,14 +95,14 @@
  * With LTO_CLANG, the linker also splits sections by default, so we need
  * these macros to combine the sections during the final link.
  *
- * With LTO_CLANG, the linker also splits sections by default, so we need
- * these macros to combine the sections during the final link.
+ * CONFIG_AUTOFD_CLANG and CONFIG_PROPELLER_CLANG will also split text sections
+ * and cluster them in the linking time.
  *
  * RODATA_MAIN is not used because existing code already defines .rodata.x
  * sections to be brought in with rodata.
  */
 #if defined(CONFIG_LD_DEAD_CODE_DATA_ELIMINATION) || defined(CONFIG_LTO_CLANG) || \
-defined(CONFIG_AUTOFDO_CLANG)
+defined(CONFIG_AUTOFDO_CLANG) || defined(CONFIG_PROPELLER_CLANG)
 #define TEXT_MAIN .text .text.[0-9a-zA-Z_]*
 #else
 #define TEXT_MAIN .text
@@ -612,7 +612,7 @@ defined(CONFIG_AUTOFDO_CLANG)
  * first when in these builds.
  */
 #if defined(CONFIG_LD_DEAD_CODE_DATA_ELIMINATION) || defined(CONFIG_LTO_CLANG) || \
-defined(CONFIG_AUTOFDO_CLANG)
+defined(CONFIG_AUTOFDO_CLANG) || defined(CONFIG_PROPELLER_CLANG)
 #define TEXT_TEXT							\
 		*(.text.asan.* .text.tsan.*)				\
 		*(.text.unknown .text.unknown.*)			\
diff --git a/scripts/Makefile.lib b/scripts/Makefile.lib
index c2cab5adaf25..e239fa709c20 100644
--- a/scripts/Makefile.lib
+++ b/scripts/Makefile.lib
@@ -219,6 +219,16 @@ _c_flags += $(if $(patsubst n%,, \
 	$(CFLAGS_AUTOFDO_CLANG))
 endif
 
+#
+# Enable Clang's Propeller build flags for a file or directory depending on
+# variables AUTOFDO_PROPELLER_obj.o and PROPELLER_PROFILE.
+#
+ifeq ($(CONFIG_PROPELLER_CLANG),y)
+_c_flags += $(if $(patsubst n%,, \
+	$(AUTOFDO_PROFILE_$(basetarget).o)$(AUTOFDO_PROFILE)$(PGO_PROFILE)$(PROPELLER_PROFILE)y), \
+	$(CFLAGS_PROPELLER_CLANG))
+endif
+
 # $(src) for including checkin headers from generated source files
 # $(obj) for including generated headers from checkin source files
 ifeq ($(KBUILD_EXTMOD),)
diff --git a/scripts/Makefile.propeller b/scripts/Makefile.propeller
new file mode 100644
index 000000000000..0c9318be5f64
--- /dev/null
+++ b/scripts/Makefile.propeller
@@ -0,0 +1,25 @@
+# SPDX-License-Identifier: GPL-2.0
+
+# Enable available and selected Clang Propeller features.
+# Propeller required debug information to embed module names in the profiles.
+CFLAGS_PROPELLER_CLANG := -fdebug-info-for-profiling
+
+ifdef CLANG_PROPELLER_PROFILE_PREFIX
+CFLAGS_PROPELLER_CLANG += -fbasic-block-sections=list=$(CLANG_PROPELLER_PROFILE_PREFIX)_cc_profile.txt -ffunction-sections
+KBUILD_LDFLAGS += --symbol-ordering-file=$(CLANG_PROPELLER_PROFILE_PREFIX)_ld_profile.txt --no-warn-symbol-ordering
+else
+CFLAGS_PROPELLER_CLANG += -fbasic-block-sections=labels
+endif
+
+ifdef CONFIG_LTO_CLANG
+ifdef CONFIG_LTO_CLANG_THIN
+ifdef CLANG_PROPELLER_PROFILE_PREFIX
+KBUILD_LDFLAGS += --lto-basic-block-sections=$(CLANG_PROPELLER_PROFILE_PREFIX)_cc_profile.txt
+else
+KBUILD_LDFLAGS += --lto-basic-block-sections=labels
+endif
+endif
+else
+endif
+
+export CFLAGS_PROPELLER_CLANG
diff --git a/tools/objtool/check.c b/tools/objtool/check.c
index 254913498c3c..7cea8ba53cf4 100644
--- a/tools/objtool/check.c
+++ b/tools/objtool/check.c
@@ -4489,6 +4489,7 @@ static int validate_ibt(struct objtool_file *file)
 		    !strcmp(sec->name, "__mcount_loc")			||
 		    !strcmp(sec->name, ".kcfi_traps")			||
 		    !strcmp(sec->name, ".llvm.call-graph-profile")	||
+		    !strcmp(sec->name, ".llvm_bb_addr_map")		||
 		    strstr(sec->name, "__patchable_function_entries"))
 			continue;
 
-- 
2.46.0.rc1.232.g9752f9e123-goog

Re: [PATCH 6/6] Add Propeller configuration for kernel build.

[Arnd Bergmann]

On Sun, Jul 28, 2024, at 22:29, Rong Xu wrote:
>  Documentation/dev-tools/index.rst     |   1 +
>  Documentation/dev-tools/propeller.rst | 188 ++++++++++++++++++++++++++
>  MAINTAINERS                           |   7 +
>  Makefile                              |   1 +
>  arch/Kconfig                          |  22 +++
>  arch/x86/Kconfig                      |   1 +
>  arch/x86/boot/compressed/Makefile     |   3 +
>  arch/x86/kernel/vmlinux.lds.S         |   4 +
>  arch/x86/platform/efi/Makefile        |   1 +
>  drivers/firmware/efi/libstub/Makefile |   2 +
>  include/asm-generic/vmlinux.lds.h     |   8 +-
>  scripts/Makefile.lib                  |  10 ++
>  scripts/Makefile.propeller            |  25 ++++
>  tools/objtool/check.c                 |   1 +

I have not looked in much detail, but I see that you need
a special case for arch/x86/boot/compressed and
drivers/firmware/efi, which makes it likely that you
need to also disable properller support for
arch/x86/purgatory/Makefile, which tends to have similar
requirements.

     Arnd

Re: [PATCH 6/6] Add Propeller configuration for kernel build.

[Masahiro Yamada]

On Mon, Jul 29, 2024 at 4:02 PM Arnd Bergmann <arnd@arndb.de> wrote:
>
> On Sun, Jul 28, 2024, at 22:29, Rong Xu wrote:
> >  Documentation/dev-tools/index.rst     |   1 +
> >  Documentation/dev-tools/propeller.rst | 188 ++++++++++++++++++++++++++
> >  MAINTAINERS                           |   7 +
> >  Makefile                              |   1 +
> >  arch/Kconfig                          |  22 +++
> >  arch/x86/Kconfig                      |   1 +
> >  arch/x86/boot/compressed/Makefile     |   3 +
> >  arch/x86/kernel/vmlinux.lds.S         |   4 +
> >  arch/x86/platform/efi/Makefile        |   1 +
> >  drivers/firmware/efi/libstub/Makefile |   2 +
> >  include/asm-generic/vmlinux.lds.h     |   8 +-
> >  scripts/Makefile.lib                  |  10 ++
> >  scripts/Makefile.propeller            |  25 ++++
> >  tools/objtool/check.c                 |   1 +
>
> I have not looked in much detail, but I see that you need
> a special case for arch/x86/boot/compressed and
> drivers/firmware/efi, which makes it likely that you
> need to also disable properller support for
> arch/x86/purgatory/Makefile, which tends to have similar
> requirements.
>
>      Arnd




I applied the following commits:

 - 9c2d1328f88adb6cbfb218163623254b96f680d3
 - 7f7f6f7ad654b326897c9f54438a06f03454bd0d



This might be another case to apply a similar approach
instead of sprinkling PROPELLER__PROFILE=n.




-- 
Best Regards
Masahiro Yamada

Re: [PATCH 0/6] Add AutoFDO and Propeller support for Clang build

[Peter Zijlstra]

On Sun, Jul 28, 2024 at 01:29:53PM -0700, Rong Xu wrote:
> Hi,
> 
> This patch series is to integrate AutoFDO and Propeller support into
> the Linux kernel. AutoFDO is a profile-guided optimization technique
> that leverages hardware sampling to enhance binary performance.
> Unlike Instrumentation-based FDO (iFDO), AutoFDO offers a user-friendly
> and straightforward application process. While iFDO generally yields
> superior profile quality and performance, our findings reveal that
> AutoFDO achieves remarkable effectiveness, bringing performance close
> to iFDO for benchmark applications. Similar to AutoFDO, Propeller too
> utilizes hardware sampling to collect profiles and apply post-link
> optimizations to improve the benchmark’s performance over and above
> AutoFDO.
> 
> Our empirical data demonstrates significant performance improvements
> with AutoFDO and Propeller, up to 10% on microbenchmarks and up to 5%
> on large warehouse-scale benchmarks. This makes a strong case for their
> inclusion as supported features in the upstream kernel.
> 
> Background
> 
> A significant fraction of fleet processing cycles (excluding idle time)
> from data center workloads are attributable to the kernel. Ware-house
> scale workloads maximize performance by optimizing the production kernel
> using iFDO (a.k.a instrumented PGO, Profile Guided Optimization).
> 
> iFDO can significantly enhance application performance but its use
> within the kernel has raised concerns. AutoFDO is a variant of FDO that
> uses the hardware’s Performance Monitoring Unit (PMU) to collect
> profiling data. While AutoFDO typically yields smaller performance
> gains than iFDO, it presents unique benefits for optimizing kernels.
> 
> AutoFDO eliminates the need for instrumented kernels, allowing a single
> optimized kernel to serve both execution and profile collection. It also
> minimizes slowdown during profile collection, potentially yielding
> higher-fidelity profiling, especially for time-sensitive code, compared
> to iFDO. Additionally, AutoFDO profiles can be obtained from production
> environments via the hardware’s PMU whereas iFDO profiles require
> carefully curated load tests that are representative of real-world
> traffic.
> 
> AutoFDO facilitates profile collection across diverse targets.
> Preliminary studies indicate significant variation in kernel hot spots
> within Google’s infrastructure, suggesting potential performance gains
> through target-specific kernel customization.
> 
> Furthermore, other advanced compiler optimization techniques, including
> ThinLTO and Propeller can be stacked on top of AutoFDO, similar to iFDO.
> ThinLTO achieves better runtime performance through whole-program
> analysis and cross module optimizations. The main difference between
> traditional LTO and ThinLTO is that the latter is scalable in time and
> memory. 

This, 

> Propeller is a profile-guided, post-link optimizer that improves
> the performance of large-scale applications compiled with LLVM. It
> operates by relinking the binary based on an additional round of runtime
> profiles, enabling precise optimizations that are not possible at
> compile time.

should be on top somewhere, not hidden away inside a giant wall of text
somewhere at the end.

Re: [PATCH 2/6] objtool: Fix unreachable instruction warnings for weak funcitons

[Peter Zijlstra]

On Sun, Jul 28, 2024 at 01:29:55PM -0700, Rong Xu wrote:
> In the presence of both weak and strong function definitions, the
> linker drops the weak symbol in favor of a strong symbol, but
> leaves the code in place. Code in ignore_unreachable_insn() has
> some heuristics to suppress the warning, but it does not work when
> -ffunction-sections is enabled.
> 
> Suppose function foo has both strong and weak definitions.
> Case 1: The strong definition has an annotated section name,
> like .init.text. Only the weak definition will be placed into
> .text.foo. But since the section has no symbols, there will be no
> "hole" in the section.
> 
> Case 2: Both sections are without an annotated section name.
> Both will be placed into .text.foo section, but there will be only one
> symbol (the strong one). If the weak code is before the strong code,
> there is no "hole" as it fails to find the right-most symbol before
> the offset.
> 
> The fix is to use the first node to compute the hole if hole.sym
> is empty. If there is no symbol in the section, the first node
> will be NULL, in which case, -1 is returned to skip the whole
> section.
> 
> Co-developed-by: Han Shen <shenhan@google.com>
> Signed-off-by: Han Shen <shenhan@google.com>
> Signed-off-by: Rong Xu <xur@google.com>
> Suggested-by: Sriraman Tallam <tmsriram@google.com>
> Suggested-by: Krzysztof Pszeniczny <kpszeniczny@google.com>
> ---
>  tools/objtool/elf.c | 13 ++++++++-----
>  1 file changed, 8 insertions(+), 5 deletions(-)
> 
> diff --git a/tools/objtool/elf.c b/tools/objtool/elf.c
> index 3d27983dc908..fa88bb254ccc 100644
> --- a/tools/objtool/elf.c
> +++ b/tools/objtool/elf.c
> @@ -224,12 +224,15 @@ int find_symbol_hole_containing(const struct section *sec, unsigned long offset)
>  	if (n)
>  		return 0; /* not a hole */
>  
> -	/* didn't find a symbol for which @offset is after it */
> -	if (!hole.sym)
> -		return 0; /* not a hole */
> +	/*
> +	 * @offset >= sym->offset + sym->len, find symbol after it.
> +	 * Use the first node in rb_tree when hole.sym is NULL.
> +	 */

	/*
	 * If we are not right of any symbol, the next symbol must be
	 * the first symbol. Either way, the next symbol -- if there is
	 * one -- provides the rightmost boundary of the hole.
	 */
	if (!hole.sym)
		n = rb_first_cached(&sec->symbol_tree);
	else
		n = rb_next(&hole.sym->node);


That tells us more of why, rather than of what. Hmm?

> +	if (hole.sym)
> +		n = rb_next(&hole.sym->node);
> +	else
> +		n = rb_first_cached(&sec->symbol_tree);
>  
> -	/* @offset >= sym->offset + sym->len, find symbol after it */
> -	n = rb_next(&hole.sym->node);
>  	if (!n)
>  		return -1; /* until end of address space */
>  
> -- 
> 2.46.0.rc1.232.g9752f9e123-goog
> 

Re: [PATCH 3/6] Change the symbols order when --ffuntion-sections is enabled

[Peter Zijlstra]

On Sun, Jul 28, 2024 at 01:29:56PM -0700, Rong Xu wrote:
> When the -ffunction-sections compiler option is enabled, each function
> is placed in a separate section named .text.function_name rather than
> putting all functions in a single .text section.
> 
> However, using -function-sections can cause problems with the
> linker script. The comments included in include/asm-generic/vmlinux.lds.h
> note these issues.:
>   “TEXT_MAIN here will match .text.fixup and .text.unlikely if dead
>    code elimination is enabled, so these sections should be converted
>    to use ".." first.”
> 
> It is unclear whether there is a straightforward method for converting
> a suffix to "..". This patch modifies the order of subsections within the
> text output section when the -ffunction-sections flag is enabled.
> Specifically, it repositions sections with certain fixed patterns (for
> example .text.unlikely) before TEXT_MAIN, ensuring that they are grouped
> and matched together.
> 
> Note that the limitation arises because the linker script employs glob
> patterns instead of regular expressions for string matching. While there
> is a method to maintain the current order using complex patterns, this
> significantly complicates the pattern and increases the likelihood of
> errors.
> 
> Co-developed-by: Han Shen <shenhan@google.com>
> Signed-off-by: Han Shen <shenhan@google.com>
> Signed-off-by: Rong Xu <xur@google.com>
> Suggested-by: Sriraman Tallam <tmsriram@google.com>
> Suggested-by: Krzysztof Pszeniczny <kpszeniczny@google.com>
> ---
>  include/asm-generic/vmlinux.lds.h | 19 ++++++++++++++++---
>  1 file changed, 16 insertions(+), 3 deletions(-)
> 
> diff --git a/include/asm-generic/vmlinux.lds.h b/include/asm-generic/vmlinux.lds.h
> index 5703526d6ebf..f3de66bda293 100644
> --- a/include/asm-generic/vmlinux.lds.h
> +++ b/include/asm-generic/vmlinux.lds.h
> @@ -582,9 +582,21 @@
>   * during second ld run in second ld pass when generating System.map
>   *
>   * TEXT_MAIN here will match .text.fixup and .text.unlikely if dead
> - * code elimination is enabled, so these sections should be converted
> - * to use ".." first.
> + * code elimination or function-section is enabled. Match these symbols
> + * first when in these builds.
>   */
> +#if defined(CONFIG_LD_DEAD_CODE_DATA_ELIMINATION) || defined(CONFIG_LTO_CLANG)
> +#define TEXT_TEXT							\
> +		*(.text.asan.* .text.tsan.*)				\
> +		*(.text.unknown .text.unknown.*)			\
> +		*(.text.unlikely .text.unlikely.*)			\
> +		ALIGN_FUNCTION();					\

Why leave the above text sections unaligned?

> +		*(.text.hot .text.hot.*)				\
> +		*(TEXT_MAIN .text.fixup)				\
> +		NOINSTR_TEXT						\
> +		*(.ref.text)						\
> +	MEM_KEEP(init.text*)
> +#else
>  #define TEXT_TEXT							\
>  		ALIGN_FUNCTION();					\
>  		*(.text.hot .text.hot.*)				\
> @@ -594,7 +606,8 @@
>  		NOINSTR_TEXT						\
>  		*(.ref.text)						\
>  		*(.text.asan.* .text.tsan.*)				\
> -	MEM_KEEP(init.text*)						\
> +	MEM_KEEP(init.text*)
> +#endif
>  
>  
>  /* sched.text is aling to function alignment to secure we have same
> -- 
> 2.46.0.rc1.232.g9752f9e123-goog
> 

Re: [PATCH 6/6] Add Propeller configuration for kernel build.

[Peter Zijlstra]

On Sun, Jul 28, 2024 at 01:29:59PM -0700, Rong Xu wrote:
> Add the build support for using Clang's Propeller optimizer. Like
> AutoFDO, Propeller uses hardware sampling to gather information
> about the frequency of execution of different code paths within a
> binary. This information is then used to guide the compiler's
> optimization decisions, resulting in a more efficient binary.
> 
> The support requires a Clang compiler LLVM 19 or later, and the
> create_llvm_prof tool
> (https://github.com/google/autofdo/releases/tag/v0.30.1). This

What's the relation between this and llvm-profgen? Is the above simply
a google 'internal' proof of concept thing that will eventually make its
way into llvm-profgen?

It seems a bit weird LLVM landed propeller without the required profile
generation tool.

Re: [PATCH 6/6] Add Propeller configuration for kernel build.

[Rong Xu]

On Mon, Jul 29, 2024 at 2:53 AM Peter Zijlstra <peterz@infradead.org> wrote:
>
> On Sun, Jul 28, 2024 at 01:29:59PM -0700, Rong Xu wrote:
> > Add the build support for using Clang's Propeller optimizer. Like
> > AutoFDO, Propeller uses hardware sampling to gather information
> > about the frequency of execution of different code paths within a
> > binary. This information is then used to guide the compiler's
> > optimization decisions, resulting in a more efficient binary.
> >
> > The support requires a Clang compiler LLVM 19 or later, and the
> > create_llvm_prof tool
> > (https://github.com/google/autofdo/releases/tag/v0.30.1). This
>
> What's the relation between this and llvm-profgen? Is the above simply
> a google 'internal' proof of concept thing that will eventually make its
> way into llvm-profgen?
>
> It seems a bit weird LLVM landed propeller without the required profile
> generation tool.

AutoFDO historically required a third-party tool
(create_llvm_prof, mentioned above) in Clang.
AutoFDO in GCC also requires this tool (same source, different name).

llvm-profgen is a relatively new tool that was integrated directly into LLVM
(similar purpose, but different implementation).

AutoFDO and create_llvm_prof are not "proof of concept". They have been widely
used in Google for many years. Propeller currently uses create_llvm_prof as the
profile converting tool. But there is an effort to move to LLVM.

Re: [PATCH 0/6] Add AutoFDO and Propeller support for Clang build

[Rong Xu]

On Mon, Jul 29, 2024 at 1:51 AM Peter Zijlstra <peterz@infradead.org> wrote:
>
> On Sun, Jul 28, 2024 at 01:29:53PM -0700, Rong Xu wrote:
> > Hi,
> >
> > This patch series is to integrate AutoFDO and Propeller support into
> > the Linux kernel. AutoFDO is a profile-guided optimization technique
> > that leverages hardware sampling to enhance binary performance.
> > Unlike Instrumentation-based FDO (iFDO), AutoFDO offers a user-friendly
> > and straightforward application process. While iFDO generally yields
> > superior profile quality and performance, our findings reveal that
> > AutoFDO achieves remarkable effectiveness, bringing performance close
> > to iFDO for benchmark applications. Similar to AutoFDO, Propeller too
> > utilizes hardware sampling to collect profiles and apply post-link
> > optimizations to improve the benchmark’s performance over and above
> > AutoFDO.
> >
> > Our empirical data demonstrates significant performance improvements
> > with AutoFDO and Propeller, up to 10% on microbenchmarks and up to 5%
> > on large warehouse-scale benchmarks. This makes a strong case for their
> > inclusion as supported features in the upstream kernel.
> >
> > Background
> >
> > A significant fraction of fleet processing cycles (excluding idle time)
> > from data center workloads are attributable to the kernel. Ware-house
> > scale workloads maximize performance by optimizing the production kernel
> > using iFDO (a.k.a instrumented PGO, Profile Guided Optimization).
> >
> > iFDO can significantly enhance application performance but its use
> > within the kernel has raised concerns. AutoFDO is a variant of FDO that
> > uses the hardware’s Performance Monitoring Unit (PMU) to collect
> > profiling data. While AutoFDO typically yields smaller performance
> > gains than iFDO, it presents unique benefits for optimizing kernels.
> >
> > AutoFDO eliminates the need for instrumented kernels, allowing a single
> > optimized kernel to serve both execution and profile collection. It also
> > minimizes slowdown during profile collection, potentially yielding
> > higher-fidelity profiling, especially for time-sensitive code, compared
> > to iFDO. Additionally, AutoFDO profiles can be obtained from production
> > environments via the hardware’s PMU whereas iFDO profiles require
> > carefully curated load tests that are representative of real-world
> > traffic.
> >
> > AutoFDO facilitates profile collection across diverse targets.
> > Preliminary studies indicate significant variation in kernel hot spots
> > within Google’s infrastructure, suggesting potential performance gains
> > through target-specific kernel customization.
> >
> > Furthermore, other advanced compiler optimization techniques, including
> > ThinLTO and Propeller can be stacked on top of AutoFDO, similar to iFDO.
> > ThinLTO achieves better runtime performance through whole-program
> > analysis and cross module optimizations. The main difference between
> > traditional LTO and ThinLTO is that the latter is scalable in time and
> > memory.
>
> This,
>
> > Propeller is a profile-guided, post-link optimizer that improves
> > the performance of large-scale applications compiled with LLVM. It
> > operates by relinking the binary based on an additional round of runtime
> > profiles, enabling precise optimizations that are not possible at
> > compile time.
>
> should be on top somewhere, not hidden away inside a giant wall of text
> somewhere at the end.

Thanks for the suggestion. I'll move it up. Maybe after the first
paragraph in Background.

Sorry if you received a duplicated message -- I'm resending this in
plain text mode.

-Rong

Re: [PATCH 3/6] Change the symbols order when --ffuntion-sections is enabled

[Rong Xu]

On Mon, Jul 29, 2024 at 2:34 AM Peter Zijlstra <peterz@infradead.org> wrote:
>
> On Sun, Jul 28, 2024 at 01:29:56PM -0700, Rong Xu wrote:
> > When the -ffunction-sections compiler option is enabled, each function
> > is placed in a separate section named .text.function_name rather than
> > putting all functions in a single .text section.
> >
> > However, using -function-sections can cause problems with the
> > linker script. The comments included in include/asm-generic/vmlinux.lds.h
> > note these issues.:
> >   “TEXT_MAIN here will match .text.fixup and .text.unlikely if dead
> >    code elimination is enabled, so these sections should be converted
> >    to use ".." first.”
> >
> > It is unclear whether there is a straightforward method for converting
> > a suffix to "..". This patch modifies the order of subsections within the
> > text output section when the -ffunction-sections flag is enabled.
> > Specifically, it repositions sections with certain fixed patterns (for
> > example .text.unlikely) before TEXT_MAIN, ensuring that they are grouped
> > and matched together.
> >
> > Note that the limitation arises because the linker script employs glob
> > patterns instead of regular expressions for string matching. While there
> > is a method to maintain the current order using complex patterns, this
> > significantly complicates the pattern and increases the likelihood of
> > errors.
> >
> > Co-developed-by: Han Shen <shenhan@google.com>
> > Signed-off-by: Han Shen <shenhan@google.com>
> > Signed-off-by: Rong Xu <xur@google.com>
> > Suggested-by: Sriraman Tallam <tmsriram@google.com>
> > Suggested-by: Krzysztof Pszeniczny <kpszeniczny@google.com>
> > ---
> >  include/asm-generic/vmlinux.lds.h | 19 ++++++++++++++++---
> >  1 file changed, 16 insertions(+), 3 deletions(-)
> >
> > diff --git a/include/asm-generic/vmlinux.lds.h b/include/asm-generic/vmlinux.lds.h
> > index 5703526d6ebf..f3de66bda293 100644
> > --- a/include/asm-generic/vmlinux.lds.h
> > +++ b/include/asm-generic/vmlinux.lds.h
> > @@ -582,9 +582,21 @@
> >   * during second ld run in second ld pass when generating System.map
> >   *
> >   * TEXT_MAIN here will match .text.fixup and .text.unlikely if dead
> > - * code elimination is enabled, so these sections should be converted
> > - * to use ".." first.
> > + * code elimination or function-section is enabled. Match these symbols
> > + * first when in these builds.
> >   */
> > +#if defined(CONFIG_LD_DEAD_CODE_DATA_ELIMINATION) || defined(CONFIG_LTO_CLANG)
> > +#define TEXT_TEXT                                                    \
> > +             *(.text.asan.* .text.tsan.*)                            \
> > +             *(.text.unknown .text.unknown.*)                        \
> > +             *(.text.unlikely .text.unlikely.*)                      \
> > +             ALIGN_FUNCTION();                                       \
>
> Why leave the above text sections unaligned?

They are considered cold text. They are not aligned before the change.
But I have no objections to making it aligned.

(Sorry if you receive a duplicated message. I'm resending this in
plain text mode.)

>
> > +             *(.text.hot .text.hot.*)                                \
> > +             *(TEXT_MAIN .text.fixup)                                \
> > +             NOINSTR_TEXT                                            \
> > +             *(.ref.text)                                            \
> > +     MEM_KEEP(init.text*)
> > +#else
> >  #define TEXT_TEXT                                                    \
> >               ALIGN_FUNCTION();                                       \
> >               *(.text.hot .text.hot.*)                                \
> > @@ -594,7 +606,8 @@
> >               NOINSTR_TEXT                                            \
> >               *(.ref.text)                                            \
> >               *(.text.asan.* .text.tsan.*)                            \
> > -     MEM_KEEP(init.text*)                                            \
> > +     MEM_KEEP(init.text*)
> > +#endif
> >
> >
> >  /* sched.text is aling to function alignment to secure we have same
> > --
> > 2.46.0.rc1.232.g9752f9e123-goog
> >

Re: [PATCH 3/6] Change the symbols order when --ffuntion-sections is enabled

[Peter Zijlstra]

On Mon, Jul 29, 2024 at 11:48:54AM -0700, Rong Xu wrote:

> > defined(CONFIG_LTO_CLANG)
> > > +#define TEXT_TEXT                                                    \
> > > +             *(.text.asan.* .text.tsan.*)                            \
> > > +             *(.text.unknown .text.unknown.*)                        \
> > > +             *(.text.unlikely .text.unlikely.*)                      \
> > > +             ALIGN_FUNCTION();                                       \
> >
> > Why leave the above text sections unaligned?
> >
> 
> They are considered cold text. They are not aligned before the change. But
> I have no objections to making it aligned.

At least x86 has hard assumptions about function alignment always being
respected -- see the most horrible games we play with
CONFIG_CALL_THUNKS.

Or is this only text parts and not actual functions in these sections?
In which case we can probably get away with not respecting the function
call alignment, although we should probably still respect the branch
alignment -- but I forgot if we made use of that :/


> >
> > > +             *(.text.hot .text.hot.*)                                \
> > > +             *(TEXT_MAIN .text.fixup)                                \
> > > +             NOINSTR_TEXT                                            \
> > > +             *(.ref.text)                                            \
> > > +     MEM_KEEP(init.text*)
> > > +#else
> > >  #define TEXT_TEXT                                                    \
> > >               ALIGN_FUNCTION();                                       \
> > >               *(.text.hot .text.hot.*)                                \
> > > @@ -594,7 +606,8 @@
> > >               NOINSTR_TEXT                                            \
> > >               *(.ref.text)                                            \
> > >               *(.text.asan.* .text.tsan.*)                            \
> > > -     MEM_KEEP(init.text*)                                            \
> > > +     MEM_KEEP(init.text*)
> > > +#endif
> > >
> > >
> > >  /* sched.text is aling to function alignment to secure we have same
> > > --
> > > 2.46.0.rc1.232.g9752f9e123-goog
> > >
> >

Re: [PATCH 3/6] Change the symbols order when --ffuntion-sections is enabled

[H. Peter Anvin]

On July 30, 2024 1:40:22 AM PDT, Peter Zijlstra <peterz@infradead.org> wrote:
>On Mon, Jul 29, 2024 at 11:48:54AM -0700, Rong Xu wrote:
>
>> > defined(CONFIG_LTO_CLANG)
>> > > +#define TEXT_TEXT                                                    \
>> > > +             *(.text.asan.* .text.tsan.*)                            \
>> > > +             *(.text.unknown .text.unknown.*)                        \
>> > > +             *(.text.unlikely .text.unlikely.*)                      \
>> > > +             ALIGN_FUNCTION();                                       \
>> >
>> > Why leave the above text sections unaligned?
>> >
>> 
>> They are considered cold text. They are not aligned before the change. But
>> I have no objections to making it aligned.
>
>At least x86 has hard assumptions about function alignment always being
>respected -- see the most horrible games we play with
>CONFIG_CALL_THUNKS.
>
>Or is this only text parts and not actual functions in these sections?
>In which case we can probably get away with not respecting the function
>call alignment, although we should probably still respect the branch
>alignment -- but I forgot if we made use of that :/
>
>
>> >
>> > > +             *(.text.hot .text.hot.*)                                \
>> > > +             *(TEXT_MAIN .text.fixup)                                \
>> > > +             NOINSTR_TEXT                                            \
>> > > +             *(.ref.text)                                            \
>> > > +     MEM_KEEP(init.text*)
>> > > +#else
>> > >  #define TEXT_TEXT                                                    \
>> > >               ALIGN_FUNCTION();                                       \
>> > >               *(.text.hot .text.hot.*)                                \
>> > > @@ -594,7 +606,8 @@
>> > >               NOINSTR_TEXT                                            \
>> > >               *(.ref.text)                                            \
>> > >               *(.text.asan.* .text.tsan.*)                            \
>> > > -     MEM_KEEP(init.text*)                                            \
>> > > +     MEM_KEEP(init.text*)
>> > > +#endif
>> > >
>> > >
>> > >  /* sched.text is aling to function alignment to secure we have same
>> > > --
>> > > 2.46.0.rc1.232.g9752f9e123-goog
>> > >
>> >

The linker should always enforce the alignment of any input section. If we don't have proper alignment either the linker is broken or we don't have the correct .balign directives in the code – which is the right way to fix this.

Re: [PATCH 6/6] Add Propeller configuration for kernel build.

[Maksim Panchenko]

On Sun, Jul 28, 2024 at 01:29:56PM -0700, Rong Xu wrote:
> Add the build support for using Clang's Propeller optimizer. Like
> AutoFDO, Propeller uses hardware sampling to gather information
> about the frequency of execution of different code paths within a
> binary. This information is then used to guide the compiler's
> optimization decisions, resulting in a more efficient binary.

Thank you for submitting the patches with the latest compiler features.

Regarding Propeller, I want to quickly mention that I plan to send a
patch to include BOLT as a profile-based post-link optimizer for the
kernel. I'd like it to be considered an alternative that is selectable
at build time.

BOLT also uses sampling, and the profile can be collected on virtually
any kernel (with some caveats).  There are no constraints on the
compiler (i.e., any version of GCC or Clang is acceptable), while Linux
perf is the only external dependency used for profile collection and
conversion. BOLT works on top of AutoFDO and LTO but can be used without
them if the user desires. The build overhead is a few seconds.

As you've heard from the LLVM discussion
(https://discourse.llvm.org/t/optimizing-the-linux-kernel-with-autofdo-including-thinlto-and-propeller)
and LPC talk (https://lpc.events/event/18/contributions/1921/), at Meta,
we've also successfully optimized the kernel and got similar results.

Again, this is a heads-up before the patch, and I would like to hear
what people think about having a binary optimizer as a user-selectable
alternative to Propeller.

Thanks,
Maksim

Re: [PATCH 6/6] Add Propeller configuration for kernel build.

[Nick Desaulniers]

On Thu, Sep 19, 2024 at 4:52 AM Maksim Panchenko <max4bolt@gmail.com> wrote:
>
> On Sun, Jul 28, 2024 at 01:29:56PM -0700, Rong Xu wrote:
> > Add the build support for using Clang's Propeller optimizer. Like
> > AutoFDO, Propeller uses hardware sampling to gather information
> > about the frequency of execution of different code paths within a
> > binary. This information is then used to guide the compiler's
> > optimization decisions, resulting in a more efficient binary.
>
> Thank you for submitting the patches with the latest compiler features.
>
> Regarding Propeller, I want to quickly mention that I plan to send a
> patch to include BOLT as a profile-based post-link optimizer for the
> kernel. I'd like it to be considered an alternative that is selectable
> at build time.
>
> BOLT also uses sampling, and the profile can be collected on virtually
> any kernel (with some caveats).  There are no constraints on the
> compiler (i.e., any version of GCC or Clang is acceptable), while Linux
> perf is the only external dependency used for profile collection and
> conversion. BOLT works on top of AutoFDO and LTO but can be used without
> them if the user desires. The build overhead is a few seconds.
>
> As you've heard from the LLVM discussion
> (https://discourse.llvm.org/t/optimizing-the-linux-kernel-with-autofdo-including-thinlto-and-propeller)
> and LPC talk (https://lpc.events/event/18/contributions/1921/), at Meta,
> we've also successfully optimized the kernel and got similar results.
>
> Again, this is a heads-up before the patch, and I would like to hear
> what people think about having a binary optimizer as a user-selectable
> alternative to Propeller.

I'd imagine that folks would be interested in running Propeller, or
BOLT, but perhaps not both.

In that sense, Kconfig has the means to express mutual exclusion.
It's perhaps worth working together to get the kconfig selection
working such that folks can play with enabling these newer toolchain
related technologies.

The next instance of the bi-weekly public Clang Built Linux meeting is
next Wednesday. (Links from https://clangbuiltlinux.github.io/)

Perhaps it's worth Rong (and Sriraman and Han) and Maksim to stop by and chat?
-- 
Thanks,
~Nick Desaulniers

Re: [PATCH 6/6] Add Propeller configuration for kernel build.

[Nathan Chancellor]

On Fri, Sep 27, 2024 at 03:45:39PM -0700, Nick Desaulniers wrote:
> On Thu, Sep 19, 2024 at 4:52 AM Maksim Panchenko <max4bolt@gmail.com> wrote:
> >
> > On Sun, Jul 28, 2024 at 01:29:56PM -0700, Rong Xu wrote:
> > > Add the build support for using Clang's Propeller optimizer. Like
> > > AutoFDO, Propeller uses hardware sampling to gather information
> > > about the frequency of execution of different code paths within a
> > > binary. This information is then used to guide the compiler's
> > > optimization decisions, resulting in a more efficient binary.
> >
> > Thank you for submitting the patches with the latest compiler features.
> >
> > Regarding Propeller, I want to quickly mention that I plan to send a
> > patch to include BOLT as a profile-based post-link optimizer for the
> > kernel. I'd like it to be considered an alternative that is selectable
> > at build time.
> >
> > BOLT also uses sampling, and the profile can be collected on virtually
> > any kernel (with some caveats).  There are no constraints on the
> > compiler (i.e., any version of GCC or Clang is acceptable), while Linux
> > perf is the only external dependency used for profile collection and
> > conversion. BOLT works on top of AutoFDO and LTO but can be used without
> > them if the user desires. The build overhead is a few seconds.
> >
> > As you've heard from the LLVM discussion
> > (https://discourse.llvm.org/t/optimizing-the-linux-kernel-with-autofdo-including-thinlto-and-propeller)
> > and LPC talk (https://lpc.events/event/18/contributions/1921/), at Meta,
> > we've also successfully optimized the kernel and got similar results.
> >
> > Again, this is a heads-up before the patch, and I would like to hear
> > what people think about having a binary optimizer as a user-selectable
> > alternative to Propeller.
> 
> I'd imagine that folks would be interested in running Propeller, or
> BOLT, but perhaps not both.
> 
> In that sense, Kconfig has the means to express mutual exclusion.
> It's perhaps worth working together to get the kconfig selection
> working such that folks can play with enabling these newer toolchain
> related technologies.

Right, I would expect this to just be a Kconfig choice with a
description like "Post link optimization" or something of the sort, like
the RANDSTRUCT or DEBUG_INFO ones. If it does make sense to do them at
the same time, they can obviously be separate.

> The next instance of the bi-weekly public Clang Built Linux meeting is
> next Wednesday. (Links from https://clangbuiltlinux.github.io/)
> 
> Perhaps it's worth Rong (and Sriraman and Han) and Maksim to stop by and chat?

I would certainly be open to discussing the plans for upstreaming these
in the meeting. I think the sessions went well in the Toolchains Track.
There were no major objections from what I could tell.

Cheers,
Nathan

Re: [PATCH 6/6] Add Propeller configuration for kernel build.

[Masahiro Yamada]

On Mon, Jul 29, 2024 at 5:31 AM Rong Xu <xur@google.com> wrote:
>
> Add the build support for using Clang's Propeller optimizer. Like
> AutoFDO, Propeller uses hardware sampling to gather information
> about the frequency of execution of different code paths within a
> binary. This information is then used to guide the compiler's
> optimization decisions, resulting in a more efficient binary.
>
> The support requires a Clang compiler LLVM 19 or later, and the
> create_llvm_prof tool
> (https://github.com/google/autofdo/releases/tag/v0.30.1). This
> submission is limited to x86 platforms that support PMU features
> like LBR on Intel machines and AMD Zen3 BRS.
>
> For Arm, we plan to send patches for SPE-based Propeller when
> AutoFDO for Arm is ready.
>
> Here is an example workflow for building an AutoFDO+Propeller
> optimized kernel:
>
> 1) Build the kernel on the HOST machine, with AutoFDO and Propeller
>    build config
>       CONFIG_AUTOFDO_CLANG=y
>       CONFIG_PROPELLER_CLANG=y
>    then
>       $ make LLVM=1 CLANG_AUTOFDO_PROFILE=<autofdo_profile>
>
> “<autofdo_profile>” is the profile collected when doing a non-Propeller
> AutoFDO build. This step builds a kernel that has the same optimization
> level as AutoFDO, plus a metadata section that records basic block
> information. This kernel image runs as fast as an AutoFDO optimized
> kernel.
>
> 2) Install the kernel on test/production machines.
>
> 3) Run the load tests. The '-c' option in perf specifies the sample
>    event period. We suggest using a suitable prime number,
>    like 500009, for this purpose.
>    For Intel platforms:
>       $ perf record -e BR_INST_RETIRED.NEAR_TAKEN:k -a -N -b -c <count> \
>         -o <perf_file> -- <loadtest>
>    For AMD platforms:
>       The supported system are: Zen3 with BRS, or Zen4 with amd_lbr_v2
>       # To see if Zen3 support LBR:
>       $ cat proc/cpuinfo | grep " brs"
>       # To see if Zen4 support LBR:
>       $ cat proc/cpuinfo | grep amd_lbr_v2
>       # If the result is yes, then collect the profile using:
>       $ perf record --pfm-events RETIRED_TAKEN_BRANCH_INSTRUCTIONS:k -a \
>         -N -b -c <count> -o <perf_file> -- <loadtest>
>
> 4) (Optional) Download the raw perf file to the HOST machine.
>
> 5) Generate Propeller profile:
>    $ create_llvm_prof --binary=<vmlinux> --profile=<perf_file> \
>      --format=propeller --propeller_output_module_name \
>      --out=<propeller_profile_prefix>_cc_profile.txt \
>      --propeller_symorder=<propeller_profile_prefix>_ld_profile.txt
>
>    “create_llvm_prof” is the profile conversion tool, and a prebuilt
>    binary for linux can be found on
>    https://github.com/google/autofdo/releases/tag/v0.30.1 (can also build
>    from source).
>
>    "<propeller_profile_prefix>" can be something like
>    "/home/user/dir/any_string".
>
>    This command generates a pair of Propeller profiles:
>    "<propeller_profile_prefix>_cc_profile.txt" and
>    "<propeller_profile_prefix>_ld_profile.txt".
>
> 6) Rebuild the kernel using the AutoFDO and Propeller profile files.
>       CONFIG_AUTOFDO_CLANG=y
>       CONFIG_PROPELLER_CLANG=y
>    and
>       $ make LLVM=1 CLANG_AUTOFDO_PROFILE=<autofdo_profile> \
>         CLANG_PROPELLER_PROFILE_PREFIX=<propeller_profile_prefix>
>
> Co-developed-by: Han Shen <shenhan@google.com>
> Signed-off-by: Han Shen <shenhan@google.com>
> Signed-off-by: Rong Xu <xur@google.com>
> Suggested-by: Sriraman Tallam <tmsriram@google.com>
> Suggested-by: Krzysztof Pszeniczny <kpszeniczny@google.com>
> Suggested-by: Nick Desaulniers <ndesaulniers@google.com>
> Suggested-by: Stephane Eranian <eranian@google.com>
> ---





> diff --git a/Makefile b/Makefile
> index 5ae30cc94a26..85a96d973f20 100644
> --- a/Makefile
> +++ b/Makefile
> @@ -1025,6 +1025,7 @@ include-$(CONFIG_KCOV)            += scripts/Makefile.kcov
>  include-$(CONFIG_RANDSTRUCT)   += scripts/Makefile.randstruct
>  include-$(CONFIG_GCC_PLUGINS)  += scripts/Makefile.gcc-plugins
>  include-$(CONFIG_AUTOFDO_CLANG)        += scripts/Makefile.autofdo
> +include-$(CONFIG_PROPELLER_CLANG)      += scripts/Makefile.propeller



Please do not ignore this comment:

https://github.com/torvalds/linux/blob/v6.11/Makefile#L1016







> +ifdef CONFIG_LTO_CLANG
> +ifdef CONFIG_LTO_CLANG_THIN
> +ifdef CLANG_PROPELLER_PROFILE_PREFIX
> +KBUILD_LDFLAGS += --lto-basic-block-sections=$(CLANG_PROPELLER_PROFILE_PREFIX)_cc_profile.txt
> +else
> +KBUILD_LDFLAGS += --lto-basic-block-sections=labels
> +endif
> +endif
> +else
> +endif


Unreadable and redundant.


ifdef CONFIG_LTO_CLANG_THIN
  ifdef CLANG_PROPELLER_PROFILE_PREFIX
    KBUILD_LDFLAGS +=
--lto-basic-block-sections=$(CLANG_PROPELLER_PROFILE_PREFIX)_cc_profile.txt
  else
    KBUILD_LDFLAGS += --lto-basic-block-sections=labels
  endif
endif










-- 
Best Regards
Masahiro Yamada

Re: [PATCH 6/6] Add Propeller configuration for kernel build.

[Nathan Chancellor]

Hi Rong,

On Mon, Sep 30, 2024 at 10:07:05AM -0700, Rong Xu wrote:
> I don't find the Clang Build Linux meeting in the link of
> https://clangbuiltlinux.github.io/. There is no Wednesday meeting in the
> upcoming event. Can you confirm there is such a meeting.
> We will be happy to join to chat about this.

It is in the "Useful links" section, under the "Bi-weekly video meeting"
line. I'll copy it here just to make sure you have it.

Calendar, which should show the October 2nd meeting at 12pm Pacific
time:
https://calendar.google.com/calendar/embed?src=47005f8f50f21da6133d7239f3cb93d1624d2e1949963ea75dd86d5f2d5721e0%40group.calendar.google.com

Meeting link:
https://meet.google.com/wrr-mxkn-hdo

Cheers,
Nathan

Re: [PATCH 6/6] Add Propeller configuration for kernel build.

[Rong Xu]

Thanks Nathan! We will be there for the discussion.We are also happy
to discuss any additional comments or suggestions about the patch.

-Rong

On Mon, Sep 30, 2024 at 1:29 PM Nathan Chancellor <nathan@kernel.org> wrote:
>
> Hi Rong,
>
> On Mon, Sep 30, 2024 at 10:07:05AM -0700, Rong Xu wrote:
> > I don't find the Clang Build Linux meeting in the link of
> > https://clangbuiltlinux.github.io/. There is no Wednesday meeting in the
> > upcoming event. Can you confirm there is such a meeting.
> > We will be happy to join to chat about this.
>
> It is in the "Useful links" section, under the "Bi-weekly video meeting"
> line. I'll copy it here just to make sure you have it.
>
> Calendar, which should show the October 2nd meeting at 12pm Pacific
> time:
> https://calendar.google.com/calendar/embed?src=47005f8f50f21da6133d7239f3cb93d1624d2e1949963ea75dd86d5f2d5721e0%40group.calendar.google.com
>
> Meeting link:
> https://meet.google.com/wrr-mxkn-hdo
>
> Cheers,
> Nathan

Re: [PATCH 6/6] Add Propeller configuration for kernel build.

[Josh Poimboeuf]

On Sun, Sep 29, 2024 at 08:08:43PM +0900, Masahiro Yamada wrote:
> > +++ b/Makefile
> > @@ -1025,6 +1025,7 @@ include-$(CONFIG_KCOV)            += scripts/Makefile.kcov
> >  include-$(CONFIG_RANDSTRUCT)   += scripts/Makefile.randstruct
> >  include-$(CONFIG_GCC_PLUGINS)  += scripts/Makefile.gcc-plugins
> >  include-$(CONFIG_AUTOFDO_CLANG)        += scripts/Makefile.autofdo
> > +include-$(CONFIG_PROPELLER_CLANG)      += scripts/Makefile.propeller
> 
> 
> 
> Please do not ignore this comment:
> 
> https://github.com/torvalds/linux/blob/v6.11/Makefile#L1016

That comment is well hidden, it really belongs right before the
gcc-plugins line.

-- 
Josh