Add hypervisor sysfs-like support
v3 cover Add hypervisor sysfs-like support
On the 2019 Xen developer summit there was agreement that the Xen hypervisor should gain support for a hierarchical name-value store similar to the Linux kernel's sysfs.
This is a first implementation of that idea adding the basic functionality to hypervisor and tools side. The interface to any user program making use of that "xen-hypfs" is a new library "libxenhypfs" with a stable interface.
As a followup switching libxl to use the hypervisor file system for runtime parameter modifications is possible. This would enable removal of the XEN_SYSCTL_set_parameter sysctl operation.
Changes in V3:
- major rework, especially by supporting binary contents of entries
- added several new patches (1, 2, 7)
- full support of all runtime parameters
- support of writing entries (especially runtime parameters)
Changes in V2:
- all comments to V1 addressed
- added man-page for xenhypfs tool
- added runtime parameter read access for string parameters
Changes in V1:
- renamed xenfs ->xenhypfs
- added writable entries support at the interface level and in the xenhypfs tool
- added runtime parameter read access (integer type only for now)
- added docs/misc/hypfs-paths.pandoc for path descriptions
Juergen Gross (9):
xen: add a generic way to include binary files as variables
xen: split parameter related definitions in own header file
docs: add feature document for Xen hypervisor sysfs-like support
xen: add basic hypervisor filesystem support
libs: add libxenhypfs
tools: add xenfs tool
xen: provide version information in hypfs
xen: add /buildinfo/config entry to hypervisor filesystem
xen: add runtime parameter access support to hypfs
.gitignore | 6 + docs/features/hypervisorfs.pandoc | 86 +++++ docs/man/xenhypfs.1.pod | 61 ++++ docs/misc/hypfs-paths.pandoc | 153 +++++++++ tools/Rules.mk | 6 + tools/libs/Makefile | 1 + tools/libs/hypfs/Makefile | 16 + tools/libs/hypfs/core.c | 535 +++++++++++++++++++++++++++++++ tools/libs/hypfs/include/xenhypfs.h | 75 +++++ tools/libs/hypfs/libxenhypfs.map | 10 + tools/libs/hypfs/xenhypfs.pc.in | 10 + tools/misc/Makefile | 6 + tools/misc/xenhypfs.c | 189 +++++++++++ xen/arch/arm/acpi/boot.c | 1 + xen/arch/arm/cpuerrata.c | 1 + xen/arch/arm/domain_build.c | 1 + xen/arch/arm/gic-v3-lpi.c | 1 + xen/arch/arm/setup.c | 1 + xen/arch/arm/smpboot.c | 1 + xen/arch/arm/traps.c | 2 + xen/arch/arm/xen.lds.S | 5 + xen/arch/x86/acpi/boot.c | 1 + xen/arch/x86/acpi/cpu_idle.c | 1 + xen/arch/x86/acpi/cpufreq/cpufreq.c | 1 + xen/arch/x86/acpi/power.c | 1 + xen/arch/x86/apic.c | 1 + xen/arch/x86/cpu/amd.c | 1 + xen/arch/x86/cpu/common.c | 1 + xen/arch/x86/cpu/mcheck/mce.c | 1 + xen/arch/x86/cpu/mcheck/mce_intel.c | 1 + xen/arch/x86/cpu/mtrr/generic.c | 1 + xen/arch/x86/cpu/mwait-idle.c | 1 + xen/arch/x86/cpu/vpmu.c | 1 + xen/arch/x86/cpuid.c | 1 + xen/arch/x86/dom0_build.c | 1 + xen/arch/x86/e820.c | 1 + xen/arch/x86/genapic/probe.c | 1 + xen/arch/x86/genapic/x2apic.c | 1 + xen/arch/x86/hpet.c | 1 + xen/arch/x86/hvm/asid.c | 1 + xen/arch/x86/hvm/hvm.c | 1 + xen/arch/x86/hvm/hypercall.c | 1 + xen/arch/x86/hvm/quirks.c | 1 + xen/arch/x86/hvm/viridian/viridian.c | 1 + xen/arch/x86/hvm/vmx/vmcs.c | 18 +- xen/arch/x86/hvm/vmx/vmx.c | 1 + xen/arch/x86/hypercall.c | 1 + xen/arch/x86/io_apic.c | 1 + xen/arch/x86/irq.c | 1 + xen/arch/x86/microcode.c | 1 + xen/arch/x86/mm.c | 1 + xen/arch/x86/mm/p2m.c | 1 + xen/arch/x86/msi.c | 1 + xen/arch/x86/nmi.c | 1 + xen/arch/x86/numa.c | 1 + xen/arch/x86/oprofile/nmi_int.c | 1 + xen/arch/x86/psr.c | 1 + xen/arch/x86/pv/domain.c | 8 +- xen/arch/x86/pv/hypercall.c | 1 + xen/arch/x86/pv/shim.c | 1 + xen/arch/x86/setup.c | 1 + xen/arch/x86/shutdown.c | 1 + xen/arch/x86/spec_ctrl.c | 1 + xen/arch/x86/tboot.c | 1 + xen/arch/x86/time.c | 1 + xen/arch/x86/traps.c | 1 + xen/arch/x86/tsx.c | 1 + xen/arch/x86/x86_64/mmconfig-shared.c | 1 + xen/arch/x86/xen.lds.S | 5 + xen/arch/x86/xstate.c | 1 + xen/common/Makefile | 13 + xen/common/argo.c | 1 + xen/common/core_parking.c | 1 + xen/common/cpupool.c | 1 + xen/common/debugtrace.c | 1 + xen/common/domain.c | 1 + xen/common/efi/boot.c | 1 + xen/common/gdbstub.c | 1 + xen/common/grant_table.c | 18 +- xen/common/hypfs.c | 403 +++++++++++++++++++++++ xen/common/kernel.c | 79 ++++- xen/common/kexec.c | 1 + xen/common/memory.c | 1 + xen/common/page_alloc.c | 1 + xen/common/rcupdate.c | 1 + xen/common/sched_credit.c | 1 + xen/common/sched_credit2.c | 1 + xen/common/schedule.c | 1 + xen/common/shutdown.c | 1 + xen/common/timer.c | 1 + xen/common/trace.c | 1 + xen/drivers/acpi/apei/hest.c | 1 + xen/drivers/acpi/tables.c | 1 + xen/drivers/char/arm-uart.c | 1 + xen/drivers/char/console.c | 50 ++- xen/drivers/char/ehci-dbgp.c | 1 + xen/drivers/char/ns16550.c | 1 + xen/drivers/char/serial.c | 1 + xen/drivers/cpufreq/cpufreq.c | 1 + xen/drivers/passthrough/amd/iommu_acpi.c | 1 + xen/drivers/passthrough/iommu.c | 1 + xen/drivers/passthrough/pci.c | 1 + xen/drivers/passthrough/vtd/dmar.c | 1 + xen/drivers/passthrough/vtd/quirks.c | 1 + xen/drivers/passthrough/vtd/x86/vtd.c | 1 + xen/drivers/passthrough/x86/ats.c | 1 + xen/drivers/video/vesa.c | 1 + xen/drivers/video/vga.c | 1 + xen/include/public/hypfs.h | 124 +++++++ xen/include/public/xen.h | 1 + xen/include/xen/hypercall.h | 8 + xen/include/xen/hypfs.h | 93 ++++++ xen/include/xen/init.h | 120 ------- xen/include/xen/kernel.h | 3 + xen/include/xen/param.h | 184 +++++++++++ xen/tools/binfile | 29 ++ xen/xsm/flask/Makefile | 5 +- xen/xsm/flask/flask-policy.S | 16 - xen/xsm/flask/flask_op.c | 1 + xen/xsm/xsm_core.c | 1 + 120 files changed, 2274 insertions(+), 151 deletions(-) create mode 100644 docs/features/hypervisorfs.pandoc create mode 100644 docs/man/xenhypfs.1.pod create mode 100644 docs/misc/hypfs-paths.pandoc create mode 100644 tools/libs/hypfs/Makefile create mode 100644 tools/libs/hypfs/core.c create mode 100644 tools/libs/hypfs/include/xenhypfs.h create mode 100644 tools/libs/hypfs/libxenhypfs.map create mode 100644 tools/libs/hypfs/xenhypfs.pc.in create mode 100644 tools/misc/xenhypfs.c create mode 100644 xen/common/hypfs.c create mode 100644 xen/include/public/hypfs.h create mode 100644 xen/include/xen/hypfs.h create mode 100644 xen/include/xen/param.h create mode 100755 xen/tools/binfile delete mode 100644 xen/xsm/flask/flask-policy.S
Signature · 7 lines
2.16.4 _______________________________________________ Xen-devel mailing list [email protected] https://lists.xenproject.org/mailman/listinfo/xen-devel
PATCH v3 1/9 xen: add a generic way to include binary files as variables
Add a new script xen/tools/binfile for including a binary file at build time being usable via a pointer and a size variable in the hypervisor.
Make use of that generic tool in xsm.
---
V3:
- new patch
--- .gitignore | 1 + xen/tools/binfile | 29 +++++++++++++++++++++++++++++ xen/xsm/flask/Makefile | 5 ++++- xen/xsm/flask/flask-policy.S | 16 ---------------- 4 files changed, 34 insertions(+), 17 deletions(-) create mode 100755 xen/tools/binfile delete mode 100644 xen/xsm/flask/flask-policy.S
.gitignore +1 -0
@@ -313,6 +313,7 @@ xen/test/livepatch/*.livepatch xen/tools/kconfig/.tmp_gtkcheck xen/tools/kconfig/.tmp_qtcheck xen/tools/symbols +xen/xsm/flask/flask-policy.S xen/xsm/flask/include/av_perm_to_string.h xen/xsm/flask/include/av_permissions.h xen/xsm/flask/include/class_to_string.h
xen/tools/binfile new file +29 -0
@@ -0,0 +1,29 @@ +#!/bin/sh +# usage: binfile [-i] <target-src.S> <binary-file> <varname> +# -i add to .init.rodata (default: .rodata) section + +[ "$1" = "-i" ] && { + shift + section=".init" +} + +target=$1 +binsource=$2 +varname=$3 + +cat <<EOF >$target +#include <asm/asm_defns.h> + + .section $section.rodata, "a", %progbits + + .global $varname +$varname: + .incbin "$binsource" +.Lend: + + .type $varname, %object + .size $varname, . - $varname + + .global ${varname}_size + ASM_INT(${varname}_size, .Lend - $varname) +EOF
xen/xsm/flask/Makefile +4 -1
@@ -30,6 +30,9 @@ $(AV_H_FILES): $(AV_H_DEPEND) obj-bin-$(CONFIG_XSM_FLASK_POLICY) += flask-policy.o flask-policy.o: policy.bin +flask-policy.S: $(XEN_ROOT)/xen/tools/binfile + $(XEN_ROOT)/xen/tools/binfile -i $@ policy.bin xsm_flask_init_policy + FLASK_BUILD_DIR := $(CURDIR) POLICY_SRC := $(FLASK_BUILD_DIR)/xenpolicy-$(XEN_FULLVERSION)
@@ -39,4 +42,4 @@ policy.bin: FORCE .PHONY: clean clean:: - rm -f $(ALL_H_FILES) *.o $(DEPS_RM) policy.* $(POLICY_SRC) + rm -f $(ALL_H_FILES) *.o $(DEPS_RM) policy.* $(POLICY_SRC) flask-policy.S
xen/xsm/flask/flask-policy.S deleted +0 -16
@@ -1,16 +0,0 @@ -#include <asm/asm_defns.h> - - .section .init.rodata, "a", %progbits - -/* const unsigned char xsm_flask_init_policy[] __initconst */ - .global xsm_flask_init_policy -xsm_flask_init_policy: - .incbin "policy.bin" -.Lend: - - .type xsm_flask_init_policy, %object - .size xsm_flask_init_policy, . - xsm_flask_init_policy - -/* const unsigned int __initconst xsm_flask_init_policy_size */ - .global xsm_flask_init_policy_size - ASM_INT(xsm_flask_init_policy_size, .Lend - xsm_flask_init_policy)
Signature · 7 lines
2.16.4 _______________________________________________ Xen-devel mailing list [email protected] https://lists.xenproject.org/mailman/listinfo/xen-devel
Re: [Xen-devel] [PATCH v3 1/9] xen: add a generic way to include binary files as variables
Quoted 18 lines · Juergen Gross
xen/tools/binfile new file +15
-0@@ -0,0 +1,29 @@ +#!/bin/sh +# usage: binfile [-i] <target-src.S> <binary-file> <varname> +# -i add to .init.rodata (default: .rodata) section + +[ "$1" = "-i" ] && { + shift + section=".init" +} + +target=$1 +binsource=$2 +varname=$3 + +cat <<EOF >$target +#include <asm/asm_defns.h>
xen/tools/binfile line 25
Quoted 10 lines
xen/tools/binfile new file +10
-0+ + .section $section.rodata, "a", %progbits + + .global $varname +$varname: + .incbin "$binsource" +.Lend: + + .type $varname, %object + .size $varname, . - $varname
I'd prefer if you used .Lend here as well.
I wonder whether, right from the beginning, there wouldn't better be a way to also request better than byte alignment for such a blob.
xen/xsm/flask/Makefile $(AV_H_FILES): $(AV_H_DEPEND) · line 34
Quoted 8 lines
xen/xsm/flask/Makefile +2
-0@@ -30,6 +30,9 @@ $(AV_H_FILES): $(AV_H_DEPEND) obj-bin-$(CONFIG_XSM_FLASK_POLICY) += flask-policy.o flask-policy.o: policy.bin +flask-policy.S: $(XEN_ROOT)/xen/tools/binfile + $(XEN_ROOT)/xen/tools/binfile -i $@ policy.bin xsm_flask_init_policy
Doesn't objcopy provide a means to convert a binary blob into an ELF object containing the binary data from the input file? If so, why involve the assembler and an intermediate file here?
Jan
_______________________________________________
Xen-devel mailing list
[email protected]
https://lists.xenproject.org/mailman/listinfo/xen-devel
Re: [Xen-devel] [PATCH v3 1/9] xen: add a generic way to include binary files as variables
Quoted 23 lines · Jan Beulich
Quoted 22 lines · Juergen Gross
xen/tools/binfile new file +19
-0@@ -0,0 +1,29 @@ +#!/bin/sh +# usage: binfile [-i] <target-src.S> <binary-file> <varname> +# -i add to .init.rodata (default: .rodata) section + +[ "$1" = "-i" ] && { + shift + section=".init" +} + +target=$1 +binsource=$2 +varname=$3 + +cat <<EOF >$target +#include <asm/asm_defns.h> + + .section $section.rodata, "a", %progbits + + .global $varname
Quoted 8 lines
xen/tools/binfile line 25
Quoted 6 lines
xen/tools/binfile new file +6
-0+$varname: + .incbin "$binsource" +.Lend: + + .type $varname, %object + .size $varname, . - $varnameI'd prefer if you used .Lend here as well.
Okay.
Quoted 3 lines
I wonder whether, right from the beginning, there wouldn't better be a way to also request better than byte alignment for such a blob.
I can add that. What about "-a <n>" for 2^n alignment?
Quoted 7 lines
Quoted 7 lines
xen/xsm/flask/Makefile +0
-0@@ -30,6 +30,9 @@ $(AV_H_FILES): $(AV_H_DEPEND) obj-bin-$(CONFIG_XSM_FLASK_POLICY) += flask-policy.o flask-policy.o: policy.bin
Quoted 6 lines
xen/xsm/flask/Makefile $(AV_H_FILES): $(AV_H_DEPEND) · line 34
Quoted 2 lines
xen/xsm/flask/Makefile +2
-0+flask-policy.S: $(XEN_ROOT)/xen/tools/binfile + $(XEN_ROOT)/xen/tools/binfile -i $@ policy.bin xsm_flask_init_policyDoesn't objcopy provide a means to convert a binary blob into an ELF object containing the binary data from the input file? If so, why involve the assembler and an intermediate file here?
I can see how to add a symbol for that purpose using a dedicated section for each blob, but how to add the size information of the blob?
In the end I just followed commit 8d5671eb31e4bf for the inclusion of the blob.
Juergen
_______________________________________________
Xen-devel mailing list
[email protected]
https://lists.xenproject.org/mailman/listinfo/xen-devel
Re: [Xen-devel] [PATCH v3 1/9] xen: add a generic way to include binary files as variables
Quoted 33 lines · Jürgen Groß
Quoted 23 lines · Jan Beulich
Quoted 22 lines · Juergen Gross
xen/tools/binfile new file +19
-0@@ -0,0 +1,29 @@ +#!/bin/sh +# usage: binfile [-i] <target-src.S> <binary-file> <varname> +# -i add to .init.rodata (default: .rodata) section + +[ "$1" = "-i" ] && { + shift + section=".init" +} + +target=$1 +binsource=$2 +varname=$3 + +cat <<EOF >$target +#include <asm/asm_defns.h> + + .section $section.rodata, "a", %progbits + + .global $varnameQuoted 8 lines
xen/tools/binfile line 25
Quoted 6 lines
xen/tools/binfile new file +6
-0+$varname: + .incbin "$binsource" +.Lend: + + .type $varname, %object + .size $varname, . - $varnameI'd prefer if you used .Lend here as well.
Quoted 7 lines
Okay.
Quoted 3 lines
I wonder whether, right from the beginning, there wouldn't better be a way to also request better than byte alignment for such a blob.
I can add that. What about "-a <n>" for 2^n alignment?
SGTM.
Quoted 13 lines
Quoted 6 lines
Quoted 6 lines
xen/xsm/flask/Makefile +0
-0@@ -30,6 +30,9 @@ $(AV_H_FILES): $(AV_H_DEPEND) obj-bin-$(CONFIG_XSM_FLASK_POLICY) += flask-policy.o flask-policy.o: policy.binQuoted 6 lines
xen/xsm/flask/Makefile $(AV_H_FILES): $(AV_H_DEPEND) · line 34
Quoted 2 lines
xen/xsm/flask/Makefile +2
-0+flask-policy.S: $(XEN_ROOT)/xen/tools/binfile + $(XEN_ROOT)/xen/tools/binfile -i $@ policy.bin xsm_flask_init_policyDoesn't objcopy provide a means to convert a binary blob into an ELF object containing the binary data from the input file? If so, why involve the assembler and an intermediate file here?
Quoted 2 lines
I can see how to add a symbol for that purpose using a dedicated section for each blob, but how to add the size information of the blob?
Hmm, right. It would be doable, but perhaps indeed not very nice. Fair enough then.
Jan
_______________________________________________
Xen-devel mailing list
[email protected]
https://lists.xenproject.org/mailman/listinfo/xen-devel
PATCH v3 2/9 xen: split parameter related definitions in own header file
In reply to Juergen GrossMove the parameter related definitions from init.h into a new header file param.h. This will avoid include hell when new dependencies are added to parameter definitions.
---
V3:
- new patch
--- xen/arch/arm/acpi/boot.c | 1 + xen/arch/arm/cpuerrata.c | 1 + xen/arch/arm/domain_build.c | 1 + xen/arch/arm/gic-v3-lpi.c | 1 + xen/arch/arm/setup.c | 1 + xen/arch/arm/smpboot.c | 1 + xen/arch/arm/traps.c | 1 + xen/arch/x86/acpi/boot.c | 1 + xen/arch/x86/acpi/cpu_idle.c | 1 + xen/arch/x86/acpi/cpufreq/cpufreq.c | 1 + xen/arch/x86/acpi/power.c | 1 + xen/arch/x86/apic.c | 1 + xen/arch/x86/cpu/amd.c | 1 + xen/arch/x86/cpu/common.c | 1 + xen/arch/x86/cpu/mcheck/mce.c | 1 + xen/arch/x86/cpu/mcheck/mce_intel.c | 1 + xen/arch/x86/cpu/mtrr/generic.c | 1 + xen/arch/x86/cpu/mwait-idle.c | 1 + xen/arch/x86/cpu/vpmu.c | 1 + xen/arch/x86/cpuid.c | 1 + xen/arch/x86/dom0_build.c | 1 + xen/arch/x86/e820.c | 1 + xen/arch/x86/genapic/probe.c | 1 + xen/arch/x86/genapic/x2apic.c | 1 + xen/arch/x86/hpet.c | 1 + xen/arch/x86/hvm/asid.c | 1 + xen/arch/x86/hvm/hvm.c | 1 + xen/arch/x86/hvm/quirks.c | 1 + xen/arch/x86/hvm/viridian/viridian.c | 1 + xen/arch/x86/hvm/vmx/vmcs.c | 1 + xen/arch/x86/hvm/vmx/vmx.c | 1 + xen/arch/x86/io_apic.c | 1 + xen/arch/x86/irq.c | 1 + xen/arch/x86/microcode.c | 1 + xen/arch/x86/mm.c | 1 + xen/arch/x86/mm/p2m.c | 1 + xen/arch/x86/msi.c | 1 + xen/arch/x86/nmi.c | 1 + xen/arch/x86/numa.c | 1 + xen/arch/x86/oprofile/nmi_int.c | 1 + xen/arch/x86/psr.c | 1 + xen/arch/x86/pv/domain.c | 1 + xen/arch/x86/pv/shim.c | 1 + xen/arch/x86/setup.c | 1 + xen/arch/x86/shutdown.c | 1 + xen/arch/x86/spec_ctrl.c | 1 + xen/arch/x86/tboot.c | 1 + xen/arch/x86/time.c | 1 + xen/arch/x86/traps.c | 1 + xen/arch/x86/tsx.c | 1 + xen/arch/x86/x86_64/mmconfig-shared.c | 1 + xen/arch/x86/xstate.c | 1 + xen/common/argo.c | 1 + xen/common/core_parking.c | 1 + xen/common/cpupool.c | 1 + xen/common/debugtrace.c | 1 + xen/common/domain.c | 1 + xen/common/efi/boot.c | 1 + xen/common/gdbstub.c | 1 + xen/common/grant_table.c | 1 + xen/common/kernel.c | 1 + xen/common/kexec.c | 1 + xen/common/memory.c | 1 + xen/common/page_alloc.c | 1 + xen/common/rcupdate.c | 1 + xen/common/sched_credit.c | 1 + xen/common/sched_credit2.c | 1 + xen/common/schedule.c | 1 + xen/common/shutdown.c | 1 + xen/common/timer.c | 1 + xen/common/trace.c | 1 + xen/drivers/acpi/apei/hest.c | 1 + xen/drivers/acpi/tables.c | 1 + xen/drivers/char/arm-uart.c | 1 + xen/drivers/char/console.c | 1 + xen/drivers/char/ehci-dbgp.c | 1 + xen/drivers/char/ns16550.c | 1 + xen/drivers/char/serial.c | 1 + xen/drivers/cpufreq/cpufreq.c | 1 + xen/drivers/passthrough/amd/iommu_acpi.c | 1 + xen/drivers/passthrough/iommu.c | 1 + xen/drivers/passthrough/pci.c | 1 + xen/drivers/passthrough/vtd/dmar.c | 1 + xen/drivers/passthrough/vtd/quirks.c | 1 + xen/drivers/passthrough/vtd/x86/vtd.c | 1 + xen/drivers/passthrough/x86/ats.c | 1 + xen/drivers/video/vesa.c | 1 + xen/drivers/video/vga.c | 1 + xen/include/xen/init.h | 120 ----------------------------- xen/include/xen/param.h | 126 +++++++++++++++++++++++++++++++ xen/xsm/flask/flask_op.c | 1 + xen/xsm/xsm_core.c | 1 + 92 files changed, 216 insertions(+), 120 deletions(-) create mode 100644 xen/include/xen/param.h
xen/arch/arm/acpi/boot.c +1 -0
@@ -30,6 +30,7 @@#include <xen/errno.h> #include <acpi/actables.h> #include <xen/mm.h> +#include <xen/param.h> #include <xen/device_tree.h> #include <asm/acpi.h>
xen/arch/arm/cpuerrata.c +1 -0
@@ -2,6 +2,7 @@#include <xen/cpumask.h> #include <xen/init.h> #include <xen/mm.h> +#include <xen/param.h> #include <xen/sizes.h> #include <xen/smp.h> #include <xen/spinlock.h>
xen/arch/arm/domain_build.c +1 -0
@@ -2,6 +2,7 @@#include <xen/compile.h> #include <xen/lib.h> #include <xen/mm.h> +#include <xen/param.h> #include <xen/domain_page.h> #include <xen/sched.h> #include <xen/sizes.h>
xen/arch/arm/gic-v3-lpi.c +1 -0
@@ -20,6 +20,7 @@#include <xen/lib.h> #include <xen/mm.h> +#include <xen/param.h> #include <xen/sched.h> #include <xen/sizes.h> #include <xen/warning.h>
xen/arch/arm/setup.c +1 -0
@@ -30,6 +30,7 @@#include <xen/init.h> #include <xen/irq.h> #include <xen/mm.h> +#include <xen/param.h> #include <xen/softirq.h> #include <xen/keyhandler.h> #include <xen/cpu.h>
xen/arch/arm/smpboot.c +1 -0
@@ -23,6 +23,7 @@#include <xen/errno.h> #include <xen/init.h> #include <xen/mm.h> +#include <xen/param.h> #include <xen/sched.h> #include <xen/smp.h> #include <xen/softirq.h>
xen/arch/arm/traps.c +1 -0
@@ -26,6 +26,7 @@#include <xen/livepatch.h> #include <xen/mem_access.h> #include <xen/mm.h> +#include <xen/param.h> #include <xen/perfc.h> #include <xen/smp.h> #include <xen/softirq.h>
xen/arch/x86/acpi/boot.c +1 -0
@@ -27,6 +27,7 @@#include <xen/acpi.h> #include <xen/irq.h> #include <xen/mm.h> +#include <xen/param.h> #include <xen/dmi.h> #include <asm/fixmap.h> #include <asm/page.h>
xen/arch/x86/acpi/cpu_idle.c +1 -0
@@ -37,6 +37,7 @@#include <xen/smp.h> #include <xen/guest_access.h> #include <xen/keyhandler.h> +#include <xen/param.h> #include <xen/trace.h> #include <xen/irq.h> #include <asm/cache.h>
xen/arch/x86/acpi/cpufreq/cpufreq.c +1 -0
@@ -31,6 +31,7 @@#include <xen/errno.h> #include <xen/delay.h> #include <xen/cpumask.h> +#include <xen/param.h> #include <xen/sched.h> #include <xen/timer.h> #include <xen/xmalloc.h>
xen/arch/x86/acpi/power.c +1 -0
@@ -14,6 +14,7 @@#include <xen/acpi.h> #include <xen/errno.h> #include <xen/iocap.h> +#include <xen/param.h> #include <xen/sched.h> #include <asm/acpi.h> #include <asm/irq.h>
xen/arch/x86/apic.c +1 -0
@@ -20,6 +20,7 @@#include <xen/errno.h> #include <xen/init.h> #include <xen/mm.h> +#include <xen/param.h> #include <xen/sched.h> #include <xen/irq.h> #include <xen/delay.h>
xen/arch/x86/cpu/amd.c +1 -0
@@ -1,6 +1,7 @@#include <xen/init.h> #include <xen/bitops.h> #include <xen/mm.h> +#include <xen/param.h> #include <xen/smp.h> #include <xen/pci.h> #include <asm/io.h>
xen/arch/x86/cpu/common.c +1 -0
@@ -1,6 +1,7 @@#include <xen/init.h> #include <xen/string.h> #include <xen/delay.h> +#include <xen/param.h> #include <xen/smp.h> #include <asm/current.h> #include <asm/debugreg.h>
xen/arch/x86/cpu/mcheck/mce.c +1 -0
@@ -6,6 +6,7 @@#include <xen/init.h> #include <xen/types.h> #include <xen/kernel.h> +#include <xen/param.h> #include <xen/smp.h> #include <xen/errno.h> #include <xen/console.h>
xen/arch/x86/cpu/mcheck/mce_intel.c +1 -0
@@ -4,6 +4,7 @@#include <xen/event.h> #include <xen/kernel.h> #include <xen/delay.h> +#include <xen/param.h> #include <xen/smp.h> #include <xen/mm.h> #include <xen/cpu.h>
xen/arch/x86/cpu/mtrr/generic.c +1 -0
@@ -3,6 +3,7 @@#include <xen/lib.h> #include <xen/init.h> #include <xen/mm.h> +#include <xen/param.h> #include <xen/stdbool.h> #include <asm/flushtlb.h> #include <asm/invpcid.h>
xen/arch/x86/cpu/mwait-idle.c +1 -0
@@ -52,6 +52,7 @@#include <xen/lib.h> #include <xen/cpu.h> #include <xen/init.h> +#include <xen/param.h> #include <xen/softirq.h> #include <xen/trace.h> #include <asm/cpuidle.h>
xen/arch/x86/cpu/vpmu.c +1 -0
@@ -22,6 +22,7 @@#include <xen/event.h> #include <xen/guest_access.h> #include <xen/cpu.h> +#include <xen/param.h> #include <asm/regs.h> #include <asm/types.h> #include <asm/msr.h>
xen/arch/x86/cpuid.c +1 -0
@@ -1,5 +1,6 @@#include <xen/init.h> #include <xen/lib.h> +#include <xen/param.h> #include <xen/sched.h> #include <xen/nospec.h> #include <asm/cpuid.h>
xen/arch/x86/dom0_build.c +1 -0
@@ -7,6 +7,7 @@#include <xen/init.h> #include <xen/iocap.h> #include <xen/libelf.h> +#include <xen/param.h> #include <xen/pfn.h> #include <xen/sched.h> #include <xen/sched-if.h>
xen/arch/x86/e820.c +1 -0
@@ -1,6 +1,7 @@#include <xen/init.h> #include <xen/lib.h> #include <xen/mm.h> +#include <xen/param.h> #include <xen/compat.h> #include <xen/dmi.h> #include <xen/pfn.h>
xen/arch/x86/genapic/probe.c +1 -0
@@ -8,6 +8,7 @@#include <xen/kernel.h> #include <xen/ctype.h> #include <xen/init.h> +#include <xen/param.h> #include <asm/cache.h> #include <asm/fixmap.h> #include <asm/mpspec.h>
xen/arch/x86/genapic/x2apic.c +1 -0
@@ -19,6 +19,7 @@#include <xen/init.h> #include <xen/cpu.h> #include <xen/cpumask.h> +#include <xen/param.h> #include <asm/apicdef.h> #include <asm/genapic.h> #include <asm/apic.h>
xen/arch/x86/hpet.c +1 -0
@@ -11,6 +11,7 @@#include <xen/softirq.h> #include <xen/irq.h> #include <xen/numa.h> +#include <xen/param.h> #include <xen/sched.h> #include <asm/fixmap.h> #include <asm/div64.h>
xen/arch/x86/hvm/asid.c +1 -0
@@ -18,6 +18,7 @@#include <xen/init.h> #include <xen/lib.h> +#include <xen/param.h> #include <xen/sched.h> #include <xen/smp.h> #include <xen/percpu.h>
xen/arch/x86/hvm/hvm.c +1 -0
@@ -35,6 +35,7 @@#include <xen/mem_access.h> #include <xen/rangeset.h> #include <xen/monitor.h> +#include <xen/param.h> #include <xen/warning.h> #include <xen/vpci.h> #include <xen/nospec.h>
xen/arch/x86/hvm/quirks.c +1 -0
@@ -19,6 +19,7 @@#include <xen/lib.h> #include <xen/dmi.h> #include <xen/bitmap.h> +#include <xen/param.h> #include <asm/hvm/support.h> s8 __read_mostly hvm_port80_allowed =-1 ;
xen/arch/x86/hvm/viridian/viridian.c +1 -0
@@ -9,6 +9,7 @@#include <xen/version.h> #include <xen/hypercall.h> #include <xen/domain_page.h> +#include <xen/param.h> #include <asm/guest_access.h> #include <asm/guest/hyperv-tlfs.h> #include <asm/paging.h>
xen/arch/x86/hvm/vmx/vmcs.c +1 -0
@@ -18,6 +18,7 @@#include <xen/init.h> #include <xen/mm.h> #include <xen/lib.h> +#include <xen/param.h> #include <xen/errno.h> #include <xen/domain_page.h> #include <xen/event.h>
xen/arch/x86/hvm/vmx/vmx.c +1 -0
@@ -17,6 +17,7 @@#include <xen/init.h> #include <xen/lib.h> +#include <xen/param.h> #include <xen/trace.h> #include <xen/sched.h> #include <xen/irq.h>
xen/arch/x86/io_apic.c +1 -0
@@ -24,6 +24,7 @@#include <xen/init.h> #include <xen/irq.h> #include <xen/delay.h> +#include <xen/param.h> #include <xen/sched.h> #include <xen/acpi.h> #include <xen/keyhandler.h>
xen/arch/x86/irq.c +1 -0
@@ -10,6 +10,7 @@#include <xen/errno.h> #include <xen/event.h> #include <xen/irq.h> +#include <xen/param.h> #include <xen/perfc.h> #include <xen/sched.h> #include <xen/keyhandler.h>
xen/arch/x86/microcode.c +1 -0
@@ -26,6 +26,7 @@#include <xen/kernel.h> #include <xen/init.h> #include <xen/notifier.h> +#include <xen/param.h> #include <xen/sched.h> #include <xen/smp.h> #include <xen/softirq.h>
xen/arch/x86/mm.c +1 -0
@@ -103,6 +103,7 @@#include <xen/kernel.h> #include <xen/lib.h> #include <xen/mm.h> +#include <xen/param.h> #include <xen/domain.h> #include <xen/sched.h> #include <xen/err.h>
xen/arch/x86/mm/p2m.c +1 -0
@@ -27,6 +27,7 @@#include <xen/iommu.h> #include <xen/vm_event.h> #include <xen/event.h> +#include <xen/param.h> #include <public/vm_event.h> #include <asm/domain.h> #include <asm/page.h>
xen/arch/x86/msi.c +1 -0
@@ -14,6 +14,7 @@#include <xen/acpi.h> #include <xen/cpu.h> #include <xen/errno.h> +#include <xen/param.h> #include <xen/pci.h> #include <xen/pci_regs.h> #include <xen/iocap.h>
xen/arch/x86/nmi.c +1 -0
@@ -16,6 +16,7 @@#include <xen/init.h> #include <xen/lib.h> #include <xen/mm.h> +#include <xen/param.h> #include <xen/irq.h> #include <xen/delay.h> #include <xen/time.h>
xen/arch/x86/numa.c +1 -0
@@ -11,6 +11,7 @@#include <xen/nodemask.h> #include <xen/numa.h> #include <xen/keyhandler.h> +#include <xen/param.h> #include <xen/time.h> #include <xen/smp.h> #include <xen/pfn.h>
xen/arch/x86/oprofile/nmi_int.c +1 -0
@@ -15,6 +15,7 @@#include <xen/types.h> #include <xen/errno.h> #include <xen/init.h> +#include <xen/param.h> #include <xen/string.h> #include <xen/delay.h> #include <xen/xenoprof.h>
xen/arch/x86/psr.c +1 -0
@@ -16,6 +16,7 @@#include <xen/cpu.h> #include <xen/err.h> #include <xen/init.h> +#include <xen/param.h> #include <xen/sched.h> #include <asm/psr.h>
xen/arch/x86/pv/domain.c +1 -0
@@ -7,6 +7,7 @@#include <xen/domain_page.h> #include <xen/errno.h> #include <xen/lib.h> +#include <xen/param.h> #include <xen/sched.h> #include <asm/cpufeature.h>
xen/arch/x86/pv/shim.c +1 -0
@@ -23,6 +23,7 @@#include <xen/hypercall.h> #include <xen/init.h> #include <xen/iocap.h> +#include <xen/param.h> #include <xen/shutdown.h> #include <xen/types.h> #include <xen/consoled.h>
xen/arch/x86/setup.c +1 -0
@@ -2,6 +2,7 @@#include <xen/lib.h> #include <xen/err.h> #include <xen/grant_table.h> +#include <xen/param.h> #include <xen/sched.h> #include <xen/domain.h> #include <xen/serial.h>
xen/arch/x86/shutdown.c +1 -0
@@ -6,6 +6,7 @@#include <xen/init.h> #include <xen/lib.h> +#include <xen/param.h> #include <xen/sched.h> #include <xen/smp.h> #include <xen/delay.h>
xen/arch/x86/spec_ctrl.c +1 -0
@@ -19,6 +19,7 @@#include <xen/errno.h> #include <xen/init.h> #include <xen/lib.h> +#include <xen/param.h> #include <xen/warning.h> #include <asm/microcode.h>
xen/arch/x86/tboot.c +1 -0
@@ -1,6 +1,7 @@#include <xen/init.h> #include <xen/types.h> #include <xen/lib.h> +#include <xen/param.h> #include <xen/sched.h> #include <xen/domain_page.h> #include <xen/iommu.h>
xen/arch/x86/time.c +1 -0
@@ -14,6 +14,7 @@#include <xen/sched.h> #include <xen/lib.h> #include <xen/init.h> +#include <xen/param.h> #include <xen/time.h> #include <xen/timer.h> #include <xen/smp.h>
xen/arch/x86/traps.c +1 -0
@@ -30,6 +30,7 @@#include <xen/err.h> #include <xen/errno.h> #include <xen/mm.h> +#include <xen/param.h> #include <xen/console.h> #include <xen/shutdown.h> #include <xen/guest_access.h>
xen/arch/x86/tsx.c +1 -0
@@ -1,4 +1,5 @@#include <xen/init.h> +#include <xen/param.h> #include <asm/msr.h> /*
xen/arch/x86/x86_64/mmconfig-shared.c +1 -0
@@ -14,6 +14,7 @@#include <xen/init.h> #include <xen/mm.h> +#include <xen/param.h> #include <xen/acpi.h> #include <xen/xmalloc.h> #include <xen/pci.h>
xen/arch/x86/xstate.c +1 -0
@@ -5,6 +5,7 @@* */ +#include <xen/param.h> #include <xen/percpu.h> #include <xen/sched.h> #include <asm/current.h>
xen/common/argo.c +1 -0
@@ -24,6 +24,7 @@#include <xen/guest_access.h> #include <xen/lib.h> #include <xen/nospec.h> +#include <xen/param.h> #include <xen/sched.h> #include <xen/time.h> #include <xsm/xsm.h>
xen/common/core_parking.c +1 -0
@@ -19,6 +19,7 @@#include <xen/cpu.h> #include <xen/init.h> #include <xen/cpumask.h> +#include <xen/param.h> #include <asm/smp.h>
xen/common/cpupool.c +1 -0
@@ -14,6 +14,7 @@#include <xen/lib.h> #include <xen/init.h> #include <xen/cpumask.h> +#include <xen/param.h> #include <xen/percpu.h> #include <xen/sched.h> #include <xen/sched-if.h>
xen/common/debugtrace.c +1 -0
@@ -11,6 +11,7 @@#include <xen/keyhandler.h> #include <xen/lib.h> #include <xen/mm.h> +#include <xen/param.h> #include <xen/percpu.h> #include <xen/serial.h> #include <xen/smp.h>
xen/common/domain.c +1 -0
@@ -9,6 +9,7 @@#include <xen/lib.h> #include <xen/ctype.h> #include <xen/err.h> +#include <xen/param.h> #include <xen/sched.h> #include <xen/sched-if.h> #include <xen/domain.h>
xen/common/efi/boot.c +1 -0
@@ -11,6 +11,7 @@#include <xen/lib.h> #include <xen/mm.h> #include <xen/multiboot.h> +#include <xen/param.h> #include <xen/pci_regs.h> #include <xen/pfn.h> #if EFI_PAGE_SIZE !=PAGE_SIZE
xen/common/gdbstub.c +1 -0
@@ -40,6 +40,7 @@#include <xen/watchdog.h> #include <asm/debugger.h> #include <xen/init.h> +#include <xen/param.h> #include <xen/smp.h> #include <xen/console.h> #include <xen/errno.h>
xen/common/grant_table.c +1 -0
@@ -28,6 +28,7 @@#include <xen/lib.h> #include <xen/sched.h> #include <xen/mm.h> +#include <xen/param.h> #include <xen/event.h> #include <xen/trace.h> #include <xen/grant_table.h>
xen/common/kernel.c +1 -0
@@ -7,6 +7,7 @@#include <xen/init.h> #include <xen/lib.h> #include <xen/errno.h> +#include <xen/param.h> #include <xen/version.h> #include <xen/sched.h> #include <xen/paging.h>
xen/common/kexec.c +1 -0
@@ -12,6 +12,7 @@#include <xen/ctype.h> #include <xen/errno.h> #include <xen/guest_access.h> +#include <xen/param.h> #include <xen/watchdog.h> #include <xen/sched.h> #include <xen/types.h>
xen/common/memory.c +1 -0
@@ -11,6 +11,7 @@#include <xen/types.h> #include <xen/lib.h> #include <xen/mm.h> +#include <xen/param.h> #include <xen/perfc.h> #include <xen/sched.h> #include <xen/event.h>
xen/common/page_alloc.c +1 -0
@@ -126,6 +126,7 @@#include <xen/sched.h> #include <xen/spinlock.h> #include <xen/mm.h> +#include <xen/param.h> #include <xen/irq.h> #include <xen/softirq.h> #include <xen/domain_page.h>
xen/common/rcupdate.c +1 -0
@@ -34,6 +34,7 @@#include <xen/types.h> #include <xen/kernel.h> #include <xen/init.h> +#include <xen/param.h> #include <xen/spinlock.h> #include <xen/smp.h> #include <xen/rcupdate.h>
xen/common/sched_credit.c +1 -0
@@ -10,6 +10,7 @@#include <xen/init.h> #include <xen/lib.h> +#include <xen/param.h> #include <xen/sched.h> #include <xen/domain.h> #include <xen/delay.h>
xen/common/sched_credit2.c +1 -0
@@ -12,6 +12,7 @@#include <xen/init.h> #include <xen/lib.h> +#include <xen/param.h> #include <xen/sched.h> #include <xen/domain.h> #include <xen/delay.h>
xen/common/schedule.c +1 -0
@@ -16,6 +16,7 @@#ifndef COMPAT #include <xen/init.h> #include <xen/lib.h> +#include <xen/param.h> #include <xen/sched.h> #include <xen/domain.h> #include <xen/delay.h>
xen/common/shutdown.c +1 -0
@@ -1,5 +1,6 @@#include <xen/init.h> #include <xen/lib.h> +#include <xen/param.h> #include <xen/sched.h> #include <xen/domain.h> #include <xen/delay.h>
xen/common/timer.c +1 -0
@@ -10,6 +10,7 @@#include <xen/errno.h> #include <xen/sched.h> #include <xen/lib.h> +#include <xen/param.h> #include <xen/smp.h> #include <xen/perfc.h> #include <xen/time.h>
xen/common/trace.c +1 -0
@@ -19,6 +19,7 @@#include <asm/types.h> #include <asm/io.h> #include <xen/lib.h> +#include <xen/param.h> #include <xen/sched.h> #include <xen/smp.h> #include <xen/trace.h>
xen/drivers/acpi/apei/hest.c +1 -0
@@ -30,6 +30,7 @@#include <xen/init.h> #include <xen/kernel.h> #include <xen/mm.h> +#include <xen/param.h> #include <xen/pfn.h> #include <acpi/acpi.h> #include <acpi/apei.h>
xen/drivers/acpi/tables.c +1 -0
@@ -24,6 +24,7 @@#include <xen/init.h> #include <xen/kernel.h> +#include <xen/param.h> #include <xen/smp.h> #include <xen/string.h> #include <xen/types.h>
xen/drivers/char/arm-uart.c +1 -0
@@ -21,6 +21,7 @@#include <asm/types.h> #include <xen/console.h> #include <xen/device_tree.h> +#include <xen/param.h> #include <xen/serial.h> #include <xen/errno.h> #include <xen/acpi.h>
xen/drivers/char/console.c +1 -0
@@ -15,6 +15,7 @@#include <xen/init.h> #include <xen/event.h> #include <xen/console.h> +#include <xen/param.h> #include <xen/serial.h> #include <xen/softirq.h> #include <xen/keyhandler.h>
xen/drivers/char/ehci-dbgp.c +1 -0
@@ -8,6 +8,7 @@#include <xen/console.h> #include <xen/delay.h> #include <xen/errno.h> +#include <xen/param.h> #include <xen/pci.h> #include <xen/serial.h> #include <asm/byteorder.h>
xen/drivers/char/ns16550.c +1 -0
@@ -11,6 +11,7 @@#include <xen/console.h> #include <xen/init.h> #include <xen/irq.h> +#include <xen/param.h> #include <xen/sched.h> #include <xen/timer.h> #include <xen/serial.h>
xen/drivers/char/serial.c +1 -0
@@ -9,6 +9,7 @@#include <xen/delay.h> #include <xen/init.h> #include <xen/mm.h> +#include <xen/param.h> #include <xen/serial.h> #include <xen/cache.h>
xen/drivers/cpufreq/cpufreq.c +1 -0
@@ -31,6 +31,7 @@#include <xen/delay.h> #include <xen/cpumask.h> #include <xen/list.h> +#include <xen/param.h> #include <xen/sched.h> #include <xen/string.h> #include <xen/timer.h>
xen/drivers/passthrough/amd/iommu_acpi.c +1 -0
@@ -19,6 +19,7 @@#include <xen/errno.h> #include <xen/acpi.h> +#include <xen/param.h> #include <asm/apicdef.h> #include <asm/io_apic.h> #include <asm/amd-iommu.h>
xen/drivers/passthrough/iommu.c +1 -0
@@ -17,6 +17,7 @@#include <xen/paging.h> #include <xen/guest_access.h> #include <xen/event.h> +#include <xen/param.h> #include <xen/softirq.h> #include <xen/keyhandler.h> #include <xsm/xsm.h>
xen/drivers/passthrough/pci.c +1 -0
@@ -22,6 +22,7 @@#include <xen/prefetch.h> #include <xen/iommu.h> #include <xen/irq.h> +#include <xen/param.h> #include <xen/vm_event.h> #include <asm/hvm/irq.h> #include <xen/delay.h>
xen/drivers/passthrough/vtd/dmar.c +1 -0
@@ -24,6 +24,7 @@#include <xen/kernel.h> #include <xen/acpi.h> #include <xen/mm.h> +#include <xen/param.h> #include <xen/xmalloc.h> #include <xen/pci.h> #include <xen/pci_regs.h>
xen/drivers/passthrough/vtd/quirks.c +1 -0
@@ -17,6 +17,7 @@*/ #include <xen/irq.h> +#include <xen/param.h> #include <xen/sched.h> #include <xen/xmalloc.h> #include <xen/domain_page.h>
xen/drivers/passthrough/vtd/x86/vtd.c +1 -0
@@ -17,6 +17,7 @@* Copyright (C) Weidong Han <[email protected]> */ +#include <xen/param.h> #include <xen/sched.h> #include <xen/softirq.h> #include <xen/domain_page.h>
xen/drivers/passthrough/x86/ats.c +1 -0
@@ -12,6 +12,7 @@* this program; If not, see <http://www.gnu.org/licenses/>. */ +#include <xen/param.h> #include <xen/sched.h> #include <xen/pci.h> #include <xen/pci_regs.h>
xen/drivers/video/vesa.c +1 -0
@@ -6,6 +6,7 @@#include <xen/init.h> #include <xen/lib.h> +#include <xen/param.h> #include <xen/xmalloc.h> #include <xen/kernel.h> #include <xen/vga.h>
xen/drivers/video/vga.c +1 -0
@@ -7,6 +7,7 @@#include <xen/init.h> #include <xen/lib.h> #include <xen/mm.h> +#include <xen/param.h> #include <xen/vga.h> #include <xen/pci.h> #include <asm/io.h>
xen/include/xen/init.h +0 -120
126 lines not on this page · raw message
xen/include/xen/param.h new file +126 -0
126 lines not on this page · raw message
xen/xsm/flask/flask_op.c +1 -0
@@ -13,6 +13,7 @@#include <xsm/xsm.h> #include <xen/guest_access.h> #include <xen/err.h> +#include <xen/param.h> #include <public/xsm/flask_op.h>
xen/xsm/xsm_core.c +1 -0
@@ -13,6 +13,7 @@#include <xen/init.h> #include <xen/errno.h> #include <xen/lib.h> +#include <xen/param.h> #include <xen/hypercall.h> #include <xsm/xsm.h>
Signature · 7 lines
2.16.4 _______________________________________________ Xen-devel mailing list [email protected] https://lists.xenproject.org/mailman/listinfo/xen-devel
Re: [Xen-devel] [PATCH v3 2/9] xen: split parameter related definitions in own header file
Hi Juergen,
Quoted 3 lines · Juergen Gross
Move the parameter related definitions from
init.hinto a new header fileparam.h. This will avoid include hell when new dependencies are added to parameter definitions.
How did you find out the list of places where the new files need to be included?
Also, was it build test it on Arm?
Cheers,
Signature · 6 lines
Julien Grall _______________________________________________ Xen-devel mailing list [email protected] https://lists.xenproject.org/mailman/listinfo/xen-devel
Re: [Xen-devel] [PATCH v3 2/9] xen: split parameter related definitions in own header file
Quoted 8 lines · Julien Grall
Hi Juergen,
Quoted 3 lines · Juergen Gross
Move the parameter related definitions from
init.hinto a new header fileparam.h. This will avoid include hell when new dependencies are added to parameter definitions.How did you find out the list of places where the new files need to be included?
I did a grep for all parameter definitions.
Quoted 2 lines
Also, was it build test it on Arm?
Yes.
Juergen
_______________________________________________
Xen-devel mailing list
[email protected]
https://lists.xenproject.org/mailman/listinfo/xen-devel
Julien Grall
Acked-byRe: [Xen-devel] [PATCH v3 2/9] xen: split parameter related definitions in own header file
Quoted 11 lines · Jürgen Groß
Quoted 8 lines · Julien Grall
Hi Juergen,
Quoted 3 lines · Juergen Gross
Move the parameter related definitions from
init.hinto a new header fileparam.h. This will avoid include hell when new dependencies are added to parameter definitions.How did you find out the list of places where the new files need to be included?
I did a grep for all parameter definitions.
That's what I thought but wanted to check :).
Quoted 4 lines
Quoted 2 lines
Also, was it build test it on Arm?
Yes.
Thanks!
For Arm + common code:
Cheers,
Signature · 6 lines
Julien Grall _______________________________________________ Xen-devel mailing list [email protected] https://lists.xenproject.org/mailman/listinfo/xen-devel
Re: [Xen-devel] [PATCH v3 2/9] xen: split parameter related definitions in own header file
In reply to Juergen GrossQuoted 6 lines · Juergen Gross
Move the parameter related definitions from
init.hinto a new header fileparam.h. This will avoid include hell when new dependencies are added to parameter definitions.Signed-off-by Juergen Gross <[email protected]>
sched-stuff:
Regards
Signature · 6 lines
Dario Faggioli, Ph.D http://about.me/dario.faggioli Virtualization Software Engineer SUSE Labs, SUSE https://www.suse.com/ ------------------------------------------------------------------- <<This happens because _I_ choose it to happen!>> (Raistlin Majere)
_______________________________________________
Xen-devel mailing list
[email protected]
https://lists.xenproject.org/mailman/listinfo/xen-devel
Re: [Xen-devel] [PATCH v3 2/9] xen: split parameter related definitions in own header file
In reply to Juergen GrossQuoted 18 lines
-----Original Message-----
From: Juergen Gross <[email protected]>
Sent: 21 January 2020 08:43
To: [email protected]
Cc: Juergen Gross <[email protected]>; Stefano Stabellini <[email protected]>; Julien Grall <[email protected]>; Volodymyr Babchuk <[email protected]>; Andrew Cooper <[email protected]>; George Dunlap <[email protected]>; Ian Jackson <[email protected]>; Jan Beulich <[email protected]>; Konrad Rzeszutek Wilk <[email protected]>; Wei Liu <[email protected]>; Roger Pau Monné <[email protected]>; Durrant, Paul <[email protected]>; Jun Nakajima <[email protected]>; Kevin Tian <[email protected]>; Lukasz Hawrylko <[email protected]>; Christopher Clark <[email protected]>; Dario Faggioli <[email protected]>; Daniel De Graaf <[email protected]>
Subject: [PATCH v3 2/9] xen: split parameter related definitions in own header file
Quoted 5 lines
Move the parameter related definitions from
init.hinto a new header fileparam.h. This will avoid include hell when new dependencies are added to parameter definitions.Signed-off-by Juergen Gross <[email protected]>
_______________________________________________
Xen-devel mailing list
[email protected]
https://lists.xenproject.org/mailman/listinfo/xen-devel
Re: [Xen-devel] [PATCH v3 2/9] xen: split parameter related definitions in own header file
In reply to Juergen GrossQuoted 5 lines · Juergen Gross
Move the parameter related definitions from
init.hinto a new header fileparam.h. This will avoid include hell when new dependencies are added to parameter definitions.Signed-off-by Juergen Gross <[email protected]>
x86:
_______________________________________________
Xen-devel mailing list
[email protected]
https://lists.xenproject.org/mailman/listinfo/xen-devel
Re: [Xen-devel] [PATCH v3 2/9] xen: split parameter related definitions in own header file
In reply to Juergen GrossQuoted 8 lines
From: Juergen Gross <[email protected]>
Sent: Tuesday, January 21, 2020 4:43 PMMove the parameter related definitions from
init.hinto a new header fileparam.h. This will avoid include hell when new dependencies are added to parameter definitions.Signed-off-by Juergen Gross <[email protected]>
_______________________________________________
Xen-devel mailing list
[email protected]
https://lists.xenproject.org/mailman/listinfo/xen-devel
Re: [Xen-devel] [PATCH v3 2/9] xen: split parameter related definitions in own header file
In reply to Juergen GrossQuoted 5 lines · Juergen Gross
Move the parameter related definitions from
init.hinto a new header fileparam.h. This will avoid include hell when new dependencies are added to parameter definitions.Signed-off-by Juergen Gross <[email protected]>
There was some re-basing necessary here, which I hope I didn't screw up. I also decided to commit this despite, from a strictly formal perspective, there still being missing acks here, based on this being an entirely mechanical change to those files, and on the grounds that further re-basing would likely turn out to be needed with further delays here.
Jan
_______________________________________________
Xen-devel mailing list
[email protected]
https://lists.xenproject.org/mailman/listinfo/xen-devel
PATCH v3 3/9 docs: add feature document for Xen hypervisor sysfs-like support
In reply to Juergen GrossOn the 2019 Xen developer summit there was agreement that the Xen hypervisor should gain support for a hierarchical name-value store similar to the Linux kernel's sysfs.
In the beginning there should only be basic support: entries can be added from the hypervisor itself only, there is a simple hypercall interface to read the data.
Add a feature document for setting the base of a discussion regarding the desired functionality and the entries to add.
---
V1:
- remove the "--" prefixes of the sub-commands of the user tool (Jan Beulich)
- rename xenfs to xenhypfs (Jan Beulich)
- add "tree" and "write" options to user tool
V2:
- move example tree to the paths description (Ian Jackson)
- specify allowed characters for keys and values (Ian Jackson)
V3:
- correct introduction (writable entries)
--- docs/features/hypervisorfs.pandoc | 86 +++++++++++++++++++++++++++++++++++ docs/misc/hypfs-paths.pandoc | 95 +++++++++++++++++++++++++++++++++++++++ 2 files changed, 181 insertions(+) create mode 100644 docs/features/hypervisorfs.pandoc create mode 100644 docs/misc/hypfs-paths.pandoc
docs/features/hypervisorfs.pandoc new file +86 -0
@@ -0,0 +1,86 @@ +% Hypervisor FS +% Revision 1 + +\clearpage + +# Basics +---------------- --------------------- + Status: **Supported** + + Architectures: all + + Components: Hypervisor, toolstack +---------------- --------------------- + +# Overview + +The Hypervisor FS is a hierarchical name-value store for reporting +information to guests, especially dom0. It is similar to the Linux +kernel's sysfs. Entries and directories are created by the hypervisor, +while the toolstack is able to use a hypercall to query the entry +values or (if allowed by the hypervisor) to modify them. + +# User details + +With: + + xenhypfs ls <path> + +the user can list the entries of a specific path of the FS. Using: + + xenhypfs cat <path> + +the content of an entry can be retrieved. Using: + + xenhypfs write <path> <string> + +a writable entry can be modified. With: + + xenhypfs tree + +the complete Hypervisor FS entry tree can be printed. + +The FS paths are documented in `docs/misc/hypfs-paths.pandoc`. + +# Technical details + +Access to the hypervisor filesystem is done via the stable new hypercall +__HYPERVISOR_filesystem_op. + +* hypercall interface specification + * `xen/include/public/filesystem.h` +* hypervisor internal files + * `xen/include/xen/filesystem.h` + * `xen/common/filesystem.c` +* `libxenhypfs` + * `tools/libs/libxenhypfs/*` +* `xenhypfs` + * `tools/misc/xenhypfs.c` +* path documentation + * `docs/misc/hypfs-paths.pandoc` + +# Testing + +Any new parameters or hardware mitigations should be verified to show up +correctly in the filesystem. + +# Areas for improvement + +* More detailed access rights +* Entries per domain and/or per cpupool + +# Known issues + +* None + +# References + +* None + +# History + +------------------------------------------------------------------------ +Date Revision Version Notes +---------- -------- -------- ------------------------------------------- +2019-10-02 1 Xen 4.13 Document written +---------- -------- -------- -------------------------------------------
docs/misc/hypfs-paths.pandoc new file +95 -0
95 lines not on this page · raw message
Signature · 7 lines
2.16.4 _______________________________________________ Xen-devel mailing list [email protected] https://lists.xenproject.org/mailman/listinfo/xen-devel
Re: [Xen-devel] [PATCH v3 3/9] docs: add feature document for Xen hypervisor sysfs-like support
Hi Juergen.
Quoted 46 lines · Juergen Gross
On the 2019 Xen developer summit there was agreement that the Xen hypervisor should gain support for a hierarchical name-value store similar to the Linux kernel's sysfs.
In the beginning there should only be basic support: entries can be added from the hypervisor itself only, there is a simple hypercall interface to read the data.
Add a feature document for setting the base of a discussion regarding the desired functionality and the entries to add.
Signed-off-by Juergen Gross <[email protected]>---
V1:
- remove the "--" prefixes of the sub-commands of the user tool (Jan Beulich)
- rename xenfs to xenhypfs (Jan Beulich)
- add "tree" and "write" options to user toolV2:
- move example tree to the paths description (Ian Jackson)
- specify allowed characters for keys and values (Ian Jackson)V3:
- correct introduction (writable entries)--- docs/features/hypervisorfs.pandoc | 86 +++++++++++++++++++++++++++++++++++ docs/misc/hypfs-paths.pandoc | 95 +++++++++++++++++++++++++++++++++++++++ 2 files changed, 181 insertions(+)create mode 100644 docs/features/hypervisorfs.pandoc create mode 100644 docs/misc/hypfs-paths.pandoc
docs/features/hypervisorfs.pandoc new file +8
-0@@ -0,0 +1,86 @@ +% Hypervisor FS +% Revision 1 + +\clearpage + +# Basics +---------------- --------------------- + Status: **Supported**
docs/features/hypervisorfs.pandoc line 18
Quoted 10 lines
docs/features/hypervisorfs.pandoc new file +10
-0+ + Architectures: all + + Components: Hypervisor, toolstack +---------------- --------------------- + +# Overview + +The Hypervisor FS is a hierarchical name-value store for reporting +information to guests, especially dom0. It is similar to the Linux
I would like to get some consitency in the formatting at least within a same file. In this case, you seem to mostly use a single space the full stop. So I think you want to use single space here too.
Quoted 57 lines
docs/features/hypervisorfs.pandoc new file +57
-0+kernel's sysfs. Entries and directories are created by the hypervisor, +while the toolstack is able to use a hypercall to query the entry +values or (if allowed by the hypervisor) to modify them. + +# User details + +With: + + xenhypfs ls <path> + +the user can list the entries of a specific path of the FS. Using: + + xenhypfs cat <path> + +the content of an entry can be retrieved. Using: + + xenhypfs write <path> <string> + +a writable entry can be modified. With: + + xenhypfs tree + +the complete Hypervisor FS entry tree can be printed. + +The FS paths are documented in `docs/misc/hypfs-paths.pandoc`. + +# Technical details + +Access to the hypervisor filesystem is done via the stable new hypercall +__HYPERVISOR_filesystem_op. + +* hypercall interface specification + * `xen/include/public/filesystem.h` +* hypervisor internal files + * `xen/include/xen/filesystem.h` + * `xen/common/filesystem.c` +* `libxenhypfs` + * `tools/libs/libxenhypfs/*` +* `xenhypfs` + * `tools/misc/xenhypfs.c` +* path documentation + * `docs/misc/hypfs-paths.pandoc` + +# Testing + +Any new parameters or hardware mitigations should be verified to show up +correctly in the filesystem. + +# Areas for improvement + +* More detailed access rights +* Entries per domain and/or per cpupool + +# Known issues + +* None +
docs/features/hypervisorfs.pandoc line 85
Quoted 10 lines
docs/features/hypervisorfs.pandoc new file +10
-0+# References + +* None + +# History + +------------------------------------------------------------------------ +Date Revision Version Notes +---------- -------- -------- ------------------------------------------- +2019-10-02 1 Xen 4.13 Document written
Does this want any update? Such as using 4.14 rather than 4.13.
Quoted 41 lines
docs/features/hypervisorfs.pandoc new file +1
-0+---------- -------- -------- -------------------------------------------docs/misc/hypfs-paths.pandoc new file +34
-0@@ -0,0 +1,95 @@ +# Xenhypfs Paths + +This document attempts to define all the paths which are available +in the Xen hypervisor file system (hypfs). + +The hypervisor file system can be accessed via the xenhypfs tool. + +## Notation + +The hypervisor file system is similar to the Linux kernel's sysfs. +In this document directories are always specified with a trailing "/". + +The following notation conventions apply: + + DIRECTORY/ + + PATH = VALUES [TAGS] + +The first syntax defines a directory. It normally contains related +entries and the general scope of the directory is described. + +The second syntax defines a file entry containing values which are +either set by the hypervisor or, if the file is writable, can be set +by the user. + +PATH can contain simple regex constructs following the Perl compatible +regexp syntax described in pcre(3) or perlre(1). + +A hypervisor file system entry name can be any 0-delimited byte string +not containing any '/' character. The names "." and ".." are reserved +for file system internal use. + +VALUES are strings and can take the following forms: +
docs/misc/hypfs-paths.pandoc line 44
Quoted 10 lines
docs/misc/hypfs-paths.pandoc new file +10
-0+* STRING -- an arbitrary 0-delimited byte string. +* INTEGER -- An integer, in decimal representation unless otherwise + noted. +* "a literal string" -- literal strings are contained within quotes. +* (VALUE | VALUE | ... ) -- a set of alternatives. Alternatives are + separated by a "|" and all the alternatives are enclosed in "(" and + ")". + +Additional TAGS may follow as a comma separated set of the following +tags enclosed in square brackets.
It may be clearer if you replace a full stop with :.
However, I am not sure what are actually the tags? Do you have a concrete example how they can be used?
docs/misc/hypfs-paths.pandoc line 49
Quoted 5 lines
docs/misc/hypfs-paths.pandoc new file +5
-0+ +* w -- Path is writable by the user. This capability is usually + limited to the control domain (e.g. dom0). +* ARM | ARM32 | X86: the path is available for the respective architecture + only.
How about Arm64? Also, if it is support by both arm64 and arm32, should we use ARM or ARM32,ARM64?
Quoted 37 lines
docs/misc/hypfs-paths.pandoc new file +36
-0+* PV -- Path is valid for PV capable hypervisors only. +* HVM -- Path is valid for HVM capable hypervisors only. +* CONFIG_* -- Path is valid only in case the hypervisor was built with + the respective config option. + +## Example + +A populated Xen hypervisor file system might look like the following example: + + / + buildinfo/ directory containing build-time data + config contents of .config file used to build Xen + cpu-bugs/ x86: directory of cpu bug information + l1tf "Vulnerable" or "Not vulnerable" + mds "Vulnerable" or "Not vulnerable" + meltdown "Vulnerable" or "Not vulnerable" + spec-store-bypass "Vulnerable" or "Not vulnerable" + spectre-v1 "Vulnerable" or "Not vulnerable" + spectre-v2 "Vulnerable" or "Not vulnerable" + mitigations/ directory of mitigation settings + bti-thunk "N/A", "RETPOLINE", "LFENCE" or "JMP" + spec-ctrl "No", "IBRS+" or IBRS-" + ibpb "No" or "Yes" + l1d-flush "No" or "Yes" + md-clear "No" or "VERW" + l1tf-barrier "No" or "Yes" + active-hvm/ directory for mitigations active in hvm doamins + msr-spec-ctrl "No" or "Yes" + rsb "No" or "Yes" + eager-fpu "No" or "Yes" + md-clear "No" or "Yes" + active-pv/ directory for mitigations active in pv doamins + msr-spec-ctrl "No" or "Yes" + rsb "No" or "Yes" + eager-fpu "No" or "Yes" + md-clear "No" or "Yes"
docs/misc/hypfs-paths.pandoc line 95
Quoted 10 lines
docs/misc/hypfs-paths.pandoc new file +10
-0+ xpti "No" or list of "dom0", "domU", "PCID on" + l1tf-shadow "No" or list of "dom0", "domU" + params/ directory with hypervisor parameter values + (boot/runtime parameters) + +## General Paths + +#### / + +The root of the hypervisor file system.
Cheers,
Signature · 6 lines
Julien Grall _______________________________________________ Xen-devel mailing list [email protected] https://lists.xenproject.org/mailman/listinfo/xen-devel
Re: [Xen-devel] [PATCH v3 3/9] docs: add feature document for Xen hypervisor sysfs-like support
Quoted 55 lines · Julien Grall
Hi Juergen.
Quoted 52 lines · Juergen Gross
On the 2019 Xen developer summit there was agreement that the Xen hypervisor should gain support for a hierarchical name-value store similar to the Linux kernel's sysfs.
In the beginning there should only be basic support: entries can be added from the hypervisor itself only, there is a simple hypercall interface to read the data.
Add a feature document for setting the base of a discussion regarding the desired functionality and the entries to add.
Signed-off-by Juergen Gross <[email protected]>---
V1:
- remove the "--" prefixes of the sub-commands of the user tool (Jan Beulich)
- rename xenfs to xenhypfs (Jan Beulich)
- add "tree" and "write" options to user toolV2:
- move example tree to the paths description (Ian Jackson)
- specify allowed characters for keys and values (Ian Jackson)V3:
- correct introduction (writable entries)
---
docs/features/hypervisorfs.pandoc | 86
+++++++++++++++++++++++++++++++++++
docs/misc/hypfs-paths.pandoc | 95
+++++++++++++++++++++++++++++++++++++++2 files changed, 181 insertions(+)create mode 100644 docs/features/hypervisorfs.pandoc create mode 100644 docs/misc/hypfs-paths.pandoc
docs/features/hypervisorfs.pandoc new file +12
-0@@ -0,0 +1,86 @@ +% Hypervisor FS +% Revision 1 + +\clearpage + +# Basics +---------------- --------------------- + Status: **Supported** + + Architectures: all + + Components: Hypervisor, toolstack
Quoted 8 lines
docs/features/hypervisorfs.pandoc line 18
Quoted 6 lines
docs/features/hypervisorfs.pandoc new file +6
-0+---------------- --------------------- + +# Overview + +The Hypervisor FS is a hierarchical name-value store for reporting +information to guests, especially dom0. It is similar to the LinuxI would like to get some consitency in the formatting at least within a same file. In this case, you seem to mostly use a single space the full stop. So I think you want to use single space here too.
Either is fine with me.
I'm going to use a single space in case no one steps up and asks for double spaces after full stops.
Quoted 62 lines
Quoted 62 lines
docs/features/hypervisorfs.pandoc new file +61
-0+kernel's sysfs. Entries and directories are created by the hypervisor, +while the toolstack is able to use a hypercall to query the entry +values or (if allowed by the hypervisor) to modify them. + +# User details + +With: + + xenhypfs ls <path> + +the user can list the entries of a specific path of the FS. Using: + + xenhypfs cat <path> + +the content of an entry can be retrieved. Using: + + xenhypfs write <path> <string> + +a writable entry can be modified. With: + + xenhypfs tree + +the complete Hypervisor FS entry tree can be printed. + +The FS paths are documented in `docs/misc/hypfs-paths.pandoc`. + +# Technical details + +Access to the hypervisor filesystem is done via the stable new hypercall +__HYPERVISOR_filesystem_op. + +* hypercall interface specification + * `xen/include/public/filesystem.h` +* hypervisor internal files + * `xen/include/xen/filesystem.h` + * `xen/common/filesystem.c` +* `libxenhypfs` + * `tools/libs/libxenhypfs/*` +* `xenhypfs` + * `tools/misc/xenhypfs.c` +* path documentation + * `docs/misc/hypfs-paths.pandoc` + +# Testing + +Any new parameters or hardware mitigations should be verified to show up +correctly in the filesystem. + +# Areas for improvement + +* More detailed access rights +* Entries per domain and/or per cpupool + +# Known issues + +* None + +# References + +* None +
Quoted 8 lines
docs/features/hypervisorfs.pandoc line 85
Quoted 6 lines
docs/features/hypervisorfs.pandoc new file +6
-0+# History + +------------------------------------------------------------------------ +Date Revision Version Notes +---------- -------- -------- ------------------------------------------- +2019-10-02 1 Xen 4.13 Document writtenDoes this want any update? Such as using 4.14 rather than 4.13.
Uh, yes.
Quoted 46 lines
Quoted 46 lines
docs/features/hypervisorfs.pandoc new file +1
-0+---------- -------- -------- -------------------------------------------docs/misc/hypfs-paths.pandoc new file +38
-0@@ -0,0 +1,95 @@ +# Xenhypfs Paths + +This document attempts to define all the paths which are available +in the Xen hypervisor file system (hypfs). + +The hypervisor file system can be accessed via the xenhypfs tool. + +## Notation + +The hypervisor file system is similar to the Linux kernel's sysfs. +In this document directories are always specified with a trailing "/". + +The following notation conventions apply: + + DIRECTORY/ + + PATH = VALUES [TAGS] + +The first syntax defines a directory. It normally contains related +entries and the general scope of the directory is described. + +The second syntax defines a file entry containing values which are +either set by the hypervisor or, if the file is writable, can be set +by the user. + +PATH can contain simple regex constructs following the Perl compatible +regexp syntax described in pcre(3) or perlre(1). + +A hypervisor file system entry name can be any 0-delimited byte string +not containing any '/' character. The names "." and ".." are reserved +for file system internal use. + +VALUES are strings and can take the following forms: + +* STRING -- an arbitrary 0-delimited byte string. +* INTEGER -- An integer, in decimal representation unless otherwise + noted. +* "a literal string" -- literal strings are contained within quotes.
Quoted 8 lines
docs/misc/hypfs-paths.pandoc line 44
Quoted 6 lines
docs/misc/hypfs-paths.pandoc new file +6
-0+* (VALUE | VALUE | ... ) -- a set of alternatives. Alternatives are + separated by a "|" and all the alternatives are enclosed in "(" and + ")". + +Additional TAGS may follow as a comma separated set of the following +tags enclosed in square brackets.It may be clearer if you replace a full stop with :.
Okay.
Quoted 2 lines
However, I am not sure what are actually the tags? Do you have a concrete example how they can be used?
I'll add this one:
/cpu-bugs/active-pv/xpti (0|1) [w,X86,PV]
Quoted 9 lines
Quoted 7 lines
docs/misc/hypfs-paths.pandoc new file +4
-0+ +* w -- Path is writable by the user. This capability is usually + limited to the control domain (e.g. dom0). +* ARM | ARM32 | X86: the path is available for the respectivearchitecture
+ only.How about Arm64? Also, if it is support by both arm64 and arm32, should we use ARM or ARM32,ARM64?
ARM64 should be added and I'd suggest to use "ARM" instead of "ARM32,ARM64".
Quoted 42 lines
Quoted 42 lines
docs/misc/hypfs-paths.pandoc new file +8
-0+* PV -- Path is valid for PV capable hypervisors only. +* HVM -- Path is valid for HVM capable hypervisors only. +* CONFIG_* -- Path is valid only in case the hypervisor was built with + the respective config option. + +## Example + +A populated Xen hypervisor file system might look like the followingexample:
+
+ /
+ buildinfo/ directory containing build-time data + config contents of .config file used to build Xen + cpu-bugs/ x86: directory of cpu bug information + l1tf "Vulnerable" or "Not vulnerable" + mds "Vulnerable" or "Not vulnerable" + meltdown "Vulnerable" or "Not vulnerable" + spec-store-bypass "Vulnerable" or "Not vulnerable" + spectre-v1 "Vulnerable" or "Not vulnerable" + spectre-v2 "Vulnerable" or "Not vulnerable" + mitigations/ directory of mitigation settings + bti-thunk "N/A", "RETPOLINE", "LFENCE" or "JMP" + spec-ctrl "No", "IBRS+" or IBRS-"
+ ibpb "No" or "Yes"
+ l1d-flush "No" or "Yes"
+ md-clear "No" or "VERW"
+ l1tf-barrier "No" or "Yes"
+ active-hvm/ directory for mitigations active in hvm doamins
+ msr-spec-ctrl "No" or "Yes"
+ rsb "No" or "Yes"
+ eager-fpu "No" or "Yes"
+ md-clear "No" or "Yes"
+ active-pv/ directory for mitigations active in pv doamins
+ msr-spec-ctrl "No" or "Yes"
+ rsb "No" or "Yes"
+ eager-fpu "No" or "Yes"
+ md-clear "No" or "Yes"
+ xpti "No" or list of "dom0", "domU", "PCID on" + l1tf-shadow "No" or list of "dom0", "domU"
Quoted 8 lines
docs/misc/hypfs-paths.pandoc line 95
Quoted 8 lines
+ params/ directory with hypervisor parameter values + (boot/runtime parameters)
docs/misc/hypfs-paths.pandoc new file +6
-0+ +## General Paths + +#### / + +The root of the hypervisor file system.
Juergen
_______________________________________________
Xen-devel mailing list
[email protected]
https://lists.xenproject.org/mailman/listinfo/xen-devel
Re: [Xen-devel] [PATCH v3 3/9] docs: add feature document for Xen hypervisor sysfs-like support
Quoted 17 lines · Jürgen Groß
Quoted 1 line · Julien Grall
However, I am not sure what are actually the tags? Do you have a concrete example how they can be used?
I'll add this one:
/cpu-bugs/active-pv/xpti (0|1) [w,X86,PV]
Quoted 9 lines
Quoted 7 lines
docs/misc/hypfs-paths.pandoc new file +4
-0+ +* w -- Path is writable by the user. This capability is usually + limited to the control domain (e.g. dom0). +* ARM | ARM32 | X86: the path is available for the respectivearchitecture
+ only.How about Arm64? Also, if it is support by both arm64 and arm32, should we use ARM or ARM32,ARM64?
Quoted 2 lines
ARM64 should be added and I'd suggest to use "ARM" instead of "ARM32,ARM64".
I am happy with that.
Cheers,
Signature · 6 lines
Julien Grall _______________________________________________ Xen-devel mailing list [email protected] https://lists.xenproject.org/mailman/listinfo/xen-devel
PATCH v3 4/9 xen: add basic hypervisor filesystem support
In reply to Juergen GrossAdd the infrastructure for the hypervisor filesystem.
This includes the hypercall interface and the base functions for entry creation, deletion and modification.
---
V1:
- rename files from filesystem.* to hypfs.*
- add dummy write entry support
- rename hypercall filesystem_op to hypfs_op
- add support for unsigned integer entries
V2:
- test new entry name to be valid
V3:
- major rework, especially by supporting binary contents of entries
- addressed all comments
--- xen/arch/arm/traps.c | 1 + xen/arch/x86/hvm/hypercall.c | 1 + xen/arch/x86/hypercall.c | 1 + xen/arch/x86/pv/hypercall.c | 1 + xen/common/Makefile | 1 + xen/common/hypfs.c | 365 +++++++++++++++++++++++++++++++++++++++++++ xen/include/public/hypfs.h | 124 +++++++++++++++ xen/include/public/xen.h | 1 + xen/include/xen/hypercall.h | 8 + xen/include/xen/hypfs.h | 89 +++++++++++ 10 files changed, 592 insertions(+) create mode 100644 xen/common/hypfs.c create mode 100644 xen/include/public/hypfs.h create mode 100644 xen/include/xen/hypfs.h
xen/arch/arm/traps.c +1 -0
@@ -1382,6 +1382,7 @@ static arm_hypercall_t arm_hypercall_table[] = {#ifdef CONFIG_ARGO HYPERCALL (argo_op,5 ), #endif + HYPERCALL(hypfs_op, 5), };#ifndef NDEBUG
xen/arch/x86/hvm/hypercall.c +1 -0
@@ -144,6 +144,7 @@ static const hypercall_table_t hvm_hypercall_table[] = {#endif HYPERCALL (xenpmu_op),COMPAT_CALL (dm_op), +HYPERCALL (hypfs_op), HYPERCALL(arch_1) };
xen/arch/x86/hypercall.c +1 -0
@@ -73,6 +73,7 @@ const hypercall_args_t hypercall_args_table[NR_hypercalls] =ARGS (hvm_op,2 ),ARGS (dm_op,3 ), #endif +ARGS (hypfs_op,5 ),ARGS (mca,1 ), ARGS(arch_1, 1), };
xen/arch/x86/pv/hypercall.c +1 -0
@@ -85,6 +85,7 @@ const hypercall_table_t pv_hypercall_table[] = {HYPERCALL (hvm_op),COMPAT_CALL (dm_op), #endif +HYPERCALL (hypfs_op),HYPERCALL (mca), HYPERCALL(arch_1), };
xen/common/Makefile +1 -0
@@ -11,6 +11,7 @@ obj-y += domain.o obj-y += event_2l.o obj-y += event_channel.o obj-y += event_fifo.o +obj-y += hypfs.o obj-$(CONFIG_CRASH_DEBUG) += gdbstub.o obj-$(CONFIG_GRANT_TABLE) += grant_table.o obj-y += guestcopy.o
xen/common/hypfs.c new file +365 -0
365 lines not on this page · raw message
xen/include/public/hypfs.h new file +124 -0
124 lines not on this page · raw message
xen/include/public/xen.h +1 -0
@@ -130,6 +130,7 @@ DEFINE_XEN_GUEST_HANDLE(xen_ulong_t);#define __HYPERVISOR_argo_op 39#define __HYPERVISOR_xenpmu_op 40#define __HYPERVISOR_dm_op 41 +#define __HYPERVISOR_hypfs_op 42/* Architecture-specific hypercall definitions. */ #define __HYPERVISOR_arch_0 48
xen/include/xen/hypercall.h +8 -0
@@ -150,6 +150,14 @@ do_dm_op( unsigned int nr_bufs, XEN_GUEST_HANDLE_PARAM(xen_dm_op_buf_t) bufs); +extern long +do_hypfs_op ( +unsigned int cmd, +XEN_GUEST_HANDLE_PARAM (void ) arg1, +unsigned long arg2, +XEN_GUEST_HANDLE_PARAM (void ) arg3, +unsigned long arg4); +#ifdef CONFIG_COMPAT extern int
xen/include/xen/hypfs.h new file +89 -0
89 lines not on this page · raw message
Signature · 7 lines
2.16.4 _______________________________________________ Xen-devel mailing list [email protected] https://lists.xenproject.org/mailman/listinfo/xen-devel
Re: [Xen-devel] [PATCH v3 4/9] xen: add basic hypervisor filesystem support
On Tue, Jan 21, 2020 at 09:43:25AM +0100, Juergen Gross wrote:
[...]
Quoted 6 lines
xen/common/hypfs.c new file +0
-0
xen/common/hypfs.c line 8
Quoted 8 lines
xen/common/hypfs.c new file +8
-0@@ -0,0 +1,365 @@ +/****************************************************************************** +* +* hypfs.c +* +* Simple sysfs-like file system for the hypervisor. +*/ + +#include <xen/lib.h>
This should come after hypfs.h.
If it has come first it probably means one of the headers below hasn't included it properly?
xen/common/hypfs.c line 14
Quoted 6 lines
xen/common/hypfs.c new file +6
-0+#include <xen/err.h> +#include <xen/guest_access.h> +#include <xen/hypercall.h> +#include <xen/hypfs.h> +#include <xen/rwlock.h> +#include <public/hypfs.h>
_______________________________________________
Xen-devel mailing list
[email protected]
https://lists.xenproject.org/mailman/listinfo/xen-devel
Re: [Xen-devel] [PATCH v3 4/9] xen: add basic hypervisor filesystem support
Quoted 19 lines · Wei Liu
On Tue, Jan 21, 2020 at 09:43:25AM +0100, Juergen Gross wrote:
[...]Quoted 6 lines
xen/common/hypfs.c new file +0
-0xen/common/hypfs.c line 8
Quoted 8 lines
xen/common/hypfs.c new file +8
-0@@ -0,0 +1,365 @@ +/****************************************************************************** +* +* hypfs.c +* +* Simple sysfs-like file system for the hypervisor. +*/ + +#include <xen/lib.h> This should come after
hypfs.h.
Quoted 2 lines
If it has come first it probably means one of the headers below hasn't included it properly?
I'll move it, as the build will still be fine.
Juergen
_______________________________________________
Xen-devel mailing list
[email protected]
https://lists.xenproject.org/mailman/listinfo/xen-devel
Re: [Xen-devel] [PATCH v3 4/9] xen: add basic hypervisor filesystem support
In reply to Juergen GrossQuoted 4 lines · Juergen Gross
--- xen/arch/arm/traps.c | 1 + xen/arch/x86/hvm/hypercall.c | 1 + xen/arch/x86/hypercall.c | 1 +
Quoted 8 lines
xen/arch/x86/pv/hypercall.c | 1 + xen/common/Makefile | 1 + xen/common/hypfs.c | 365 +++++++++++++++++++++++++++++++++++++++++++ xen/include/public/hypfs.h | 124 +++++++++++++++ xen/include/public/xen.h | 1 + xen/include/xen/hypercall.h | 8 + xen/include/xen/hypfs.h | 89 +++++++++++ 10 files changed, 592 insertions(+)
Even if it's just two structures that you have in the public header, your assertion of the interface being guest bitness agnostic should be accompanied by actual proof, i.e. addition to xen/include/xlat.lst.
Quoted 25 lines
xen/common/hypfs.c new file +25
-0+static int add_entry (struct hypfs_entry_dir *parent,struct hypfs_entry *new) +{ +int ret = -ENOENT ; +struct hypfs_entry *e; + +write_lock (&hypfs_lock); + + list_for_each_entry ( e, &parent->dirlist, list ) + { +int cmp =strcmp (e->name, new->name); + +if ( cmp >0 ) + { + ret =0 ; +list_add_tail (&new->list, &e->list); +break ; + } +if ( cmp ==0 ) + { + ret = -EEXIST ; +break ; + } + } + +if ( ret == -ENOENT )
xen/common/hypfs.c line 58
Quoted 10 lines
xen/common/hypfs.c new file +10
-0+ { + ret =0 ; +list_add_tail (&new->list, &parent->dirlist); + } + +if ( !ret ) + { +unsigned int sz =strlen (new->name) +1 ; + + parent->e.size +=DIRENTRY_SIZE (sz);
Would DIRENTRY_SIZE() perhaps better include the "+ 1"?
xen/common/hypfs.c line 75
Quoted 10 lines
xen/common/hypfs.c new file +10
-0+int hypfs_add_entry (struct hypfs_entry_dir *parent, +struct hypfs_entry *entry,bool nofault) +{ +int ret; + + ret =add_entry (parent, entry); +BUG_ON (nofault && ret); + +return ret; +}
While this and its two siblings have no caller, the one above doesn't even have a declaration in the header file. What is this intended to be used for (from external callers)?
I also think the "nofault" aspect could do with discussing in the commit message - it seems quite odd to me.
xen/common/hypfs.c line 103
Quoted 2 lines
xen/common/hypfs.c new file +2
-0+static int hypfs_get_path_user (char *buf,XEN_GUEST_HANDLE_PARAM (void ) uaddr, +unsigned long len)
For consistency with naming elsewhere as well as uaddr here - ulen?
xen/common/hypfs.c line 128
Quoted 12 lines
xen/common/hypfs.c new file +12
-0+static struct hypfs_entry *hypfs_get_entry_rel (struct hypfs_entry_dir *dir, +const char *path) +{ +const char *end; +struct hypfs_entry *entry; +unsigned int name_len; + +if ( !*path ) +return &dir->e; + +if ( dir->e.type !=XEN_HYPFS_TYPE_DIR ) +return NULL ;
Don't these two need switching around, to make sure /a/b/c/ doesn't match a non-dir /a/b/c ?
xen/common/hypfs.c line 138
Quoted 9 lines
xen/common/hypfs.c new file +9
-0+ end =strchr (path,'/' ); +if ( !end ) + end =strchr (path,'\0' ); + name_len = end - path; + + list_for_each_entry ( entry, &dir->dirlist, list ) + { +int cmp =strncmp (path, entry->name, name_len); + struct hypfs_entry_dir *d = container_of(entry,
A hard tab slipped in here.
Quoted 14 lines
xen/common/hypfs.c new file +14
-0+ struct hypfs_entry_dir, e); + +if ( cmp <0 ) +return NULL ; +if ( !cmp &&strlen (entry->name) == name_len ) +return *end ?hypfs_get_entry_rel (d, end +1 ) : entry; + } + +return NULL ; +} + +struct hypfs_entry *hypfs_get_entry (const char *path) +{ +if ( path[0 ] !='/' )
xen/common/hypfs.c line 162
Quoted 10 lines
xen/common/hypfs.c new file +10
-0+return NULL ; + +return hypfs_get_entry_rel (&hypfs_root, path +1 ); +} + +int hypfs_read_dir (const struct hypfs_entry *entry, +XEN_GUEST_HANDLE_PARAM (void ) uaddr) +{ +const struct hypfs_entry_dir *d; +struct hypfs_entry *e;
const?
Quoted 11 lines
xen/common/hypfs.c new file +11
-0+static int hypfs_read (const struct hypfs_entry *entry, +XEN_GUEST_HANDLE_PARAM (void ) uaddr,unsigned long ulen) +{ +struct xen_hypfs_direntry e; +long ret = -EINVAL ; + +if ( ulen <sizeof (e) ) + goto out; + + e.flags = entry->write ?XEN_HYPFS_WRITEABLE :0 ; + e.type = entry->type;
xen/common/hypfs.c line 224
Quoted 10 lines
xen/common/hypfs.c new file +10
-0+ e.encoding = entry->encoding; + e.content_len = entry->size; + + ret = -EFAULT ; +if (copy_to_guest (uaddr, &e,1 ) ) + goto out; + + ret =0 ; +if ( ulen < entry->size +sizeof (e) ) + goto out;
So you return "success" even if the operation didn't complete successfully. This isn't very nice, plus ...
xen/common/hypfs.c line 228
Quoted 3 lines
xen/common/hypfs.c new file +3
-0+guest_handle_add_offset (uaddr,sizeof (e)); + + ret = entry->read (entry, uaddr);
... how is the caller to know whether direntry was at least copied if this then fails?
xen/common/hypfs.c line 241
Quoted 8 lines
xen/common/hypfs.c new file +8
-0+int hypfs_write_leaf (struct hypfs_entry_leaf *leaf, +XEN_GUEST_HANDLE_PARAM (void ) uaddr,unsigned long ulen) +{ +char *buf; +int ret; + +if ( ulen > leaf->e.size ) + ulen = leaf->e.size;
Silent truncation?
Quoted 1 line
+ buf =
xzalloc_array(char, ulen);
Why the z variant?
xen/common/hypfs.c line 253
Quoted 10 lines
xen/common/hypfs.c new file +10
-0+if ( !buf ) +return -ENOMEM ; + + ret = -EFAULT ; +if (copy_from_guest (buf, uaddr, ulen) ) + goto out; + + ret =0 ; +if ( leaf->e.type ==XEN_HYPFS_TYPE_STRING ) + buf[leaf->e.size -1 ] =0 ;
And possible further silent truncation? And if the incoming buffer has no nul byte at ulen-1, you'll then ...
Quoted 1 line
+
memcpy(leaf->write_ptr, buf, ulen);
... produce a strange concatenation of new and tail of old contents?
Anyway, this and ...
xen/common/hypfs.c line 263
Quoted 8 lines
+ out:
xen/common/hypfs.c new file +7
-0+xfree (buf); +return ret; +} + +int hypfs_write_bool (struct hypfs_entry_leaf *leaf, +XEN_GUEST_HANDLE_PARAM (void ) uaddr,unsigned long ulen) +{
... this function aren't very helpful to review without there being a caller. Could these be introduced at the time a first caller appears?
xen/common/hypfs.c line 275
Quoted 12 lines
xen/common/hypfs.c new file +12
-0+union { +char buf[8 ]; +uint8_t u8; +uint16_t u16; +uint32_t u32; +uint64_t u64; + } u; + +ASSERT (leaf->e.type ==XEN_HYPFS_TYPE_UINT && leaf->e.size <=8 ); + +if ( ulen != leaf->e.size ) +return -EDOM ;
Is this restriction really necessary? Setting e.g. a 4-byte field from 1-byte input is no problem at all. This being for booleans I anyway wonder why input might be helpful to have larger than a single byte. But maybe all of this is again a result of not seeing what a user of the function would look like.
Quoted 33 lines
xen/common/hypfs.c new file +33
-0+long do_hypfs_op (unsigned int cmd, +XEN_GUEST_HANDLE_PARAM (void ) arg1,unsigned long arg2, +XEN_GUEST_HANDLE_PARAM (void ) arg3,unsigned long arg4) +{ +int ret; +struct hypfs_entry *entry; +static char path[XEN_HYPFS_MAX_PATHLEN ]; + +if ( !is_control_domain (current->domain) && + !is_hardware_domain (current->domain) ) +return -EPERM ; + +if ( cmd == XEN_HYPFS_OP_get_version ) +return XEN_HYPFS_VERSION ; + +if ( cmd == XEN_HYPFS_OP_write_contents ) +write_lock (&hypfs_lock); +else +read_lock (&hypfs_lock); + + ret =hypfs_get_path_user (path, arg1, arg2); +if ( ret ) + goto out; + + entry =hypfs_get_entry (path); +if ( !entry ) + { + ret = -ENOENT ; + goto out; + } + +switch ( cmd ) + {
xen/common/hypfs.c line 354
Quoted 10 lines
xen/common/hypfs.c new file +10
-0+case XEN_HYPFS_OP_read_contents: + ret =hypfs_read (entry, arg3, arg4); +break ; + +case XEN_HYPFS_OP_write_contents: + ret =hypfs_write (entry, arg3, arg4); +break ; + +default : + ret = -ENOSYS ;
EINVAL or EOPNOTSUPP please. ENOSYS should be used only for not implemented top level functions.
Quoted 29 lines
xen/include/public/hypfs.h new file +26
-0@@ -0,0 +1,124 @@ +/****************************************************************************** +* Xen Hypervisor Filesystem +* +* Copyright (c) 2019, SUSE Software Solutions Germany GmbH +* +* Permission is hereby granted, free of charge, to any person obtaining a copy +* of this software and associated documentation files (the "Software"), to +* deal in the Software without restriction, including without limitation the +* rights to use, copy, modify, merge, publish, distribute, sublicense, and/or +* sell copies of the Software, and to permit persons to whom the Software is +* furnished to do so, subject to the following conditions: +* +* The above copyright notice and this permission notice shall be included in +* all copies or substantial portions of the Software. +* +* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR +* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, +* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE +* AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER +* LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING +* FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER +* DEALINGS IN THE SOFTWARE. +* +*/ + +#ifndef __XEN_PUBLIC_HYPFS_H__
xen/include/public/hypfs.h line 36
Quoted 10 lines
xen/include/public/hypfs.h new file +10
-0+#define __XEN_PUBLIC_HYPFS_H__ + +#include "xen.h" + +/* +* Definitions for the __HYPERVISOR_hypfs_op hypercall. +*/ + +/* Highest version number of the hypfs interface currently defined. */ +#define XEN_HYPFS_VERSION 1
For this and the accompanying XEN_HYPFS_OP_get_version, at least the doc added by patch 3 could actually do with mentioning the intentions you have with this.
Quoted 6 lines
xen/include/public/hypfs.h new file +6
-0+/* Maximum length of a path in the filesystem. */ +#define XEN_HYPFS_MAX_PATHLEN 1024 + +struct xen_hypfs_direntry { +uint16_t flags; +#define XEN_HYPFS_WRITEABLE 0x0001
xen/include/public/hypfs.h line 53
Quoted 10 lines
xen/include/public/hypfs.h new file +10
-0+uint8_t type; +#define XEN_HYPFS_TYPE_DIR 0x0000 +#define XEN_HYPFS_TYPE_BLOB 0x0001 +#define XEN_HYPFS_TYPE_STRING 0x0002 +#define XEN_HYPFS_TYPE_UINT 0x0003 +#define XEN_HYPFS_TYPE_INT 0x0004 +#define XEN_HYPFS_TYPE_BOOL 0x0005 +uint8_t encoding; +#define XEN_HYPFS_ENC_PLAIN 0x0000 +#define XEN_HYPFS_ENC_GZIP 0x0001
Meaning I can e.g. have a gzip-ed string or bool (or even dir)? If this is just for "blob", why have separate fields instead of e.g. BLOB_RAW and BLOB_GZIP or some such?
Also - why 4 digits in the constants when the fields are uint8_t? Since these are enum-like, I even wonder if hex is warranted here.
xen/include/public/hypfs.h line 62
Quoted 9 lines
xen/include/public/hypfs.h new file +9
-0+uint32_t content_len; +}; + +struct xen_hypfs_dirlistentry { +struct xen_hypfs_direntry e; +/* Offset in bytes to next entry (0 == this is the last entry). */ +uint16_t off_next; +char name[XEN_FLEX_ARRAY_DIM ]; +};
The interaction of the last two fields may want spelling out:
I _assume_ name[] is nul-terminated, and off_next exists to potentially skip trailing padding?
Quoted 18 lines
xen/include/public/hypfs.h new file +18
-0+/* +* Hypercall operations. +*/ + +/* +* XEN_HYPFS_OP_get_version +* +* Read highest interface version supported by the hypervisor. +* +* Possible return values: +* >0: highest supported interface version +* <0: negative Xen errno value +*/ +#define XEN_HYPFS_OP_get_version 0 + +/* +* XEN_HYPFS_OP_read_contents +*
xen/include/public/hypfs.h line 91
Quoted 10 lines
xen/include/public/hypfs.h new file +10
-0+* Read contents of a filesystem entry. +* +* Returns the direntry and contents of an entry in the buffer supplied by the +* caller (struct xen_hypfs_direntry with the contents following directly +* after it). +* The data buffer must be at least the size of the direntry returned in order +* to have success. If the data buffer was not large enough for all the data +* no entry data is returned, but the direntry will contain the needed size +* for the returned data. +* The format of the contents is according to its entry type and encoding.
What specifically this means for a dir would be nice to be spelled out.
Quoted 20 lines
xen/include/public/hypfs.h new file +20
-0+* arg1: XEN_GUEST_HANDLE(path name) +* arg2: length of path name (including trailing zero byte) +* arg3: XEN_GUEST_HANDLE(data buffer written by hypervisor) +* arg4: data buffer size +* +* Possible return values: +* 0: success (at least the direntry was returned) +* <0 : negative Xen errno value +*/ +#define XEN_HYPFS_OP_read_contents 1 + +/* +* XEN_HYPFS_OP_write_contents +* +* Write contents of a filesystem entry. +* +* Writes an entry with the contents of a buffer supplied by the caller. +* The data type and encoding can't be changed. The size can be changed only +* for blobs and strings. +*
xen/include/public/hypfs.h line 122
Quoted 10 lines
xen/include/public/hypfs.h new file +10
-0+* arg1: XEN_GUEST_HANDLE(path name) +* arg2: length of path name (including trailing zero byte) +* arg3: XEN_GUEST_HANDLE(content buffer read by hypervisor) +* arg4: content buffer size +* +* Possible return values: +* 0: success +* <0 : negative Xen errno value +*/ +#define XEN_HYPFS_OP_write_contents 2
This one indeed accesses only the actual data (contents) of the referenced entry. XEN_HYPFS_OP_read_contents hands back also the dir entry. Should the latter then better be named just XEN_HYPFS_OP_read?
xen/include/xen/hypercall.h do_dm_op( · line 156
Quoted 10 lines
xen/include/xen/hypercall.h +4
-0@@ -150,6 +150,14 @@ do_dm_op( unsigned int nr_bufs, XEN_GUEST_HANDLE_PARAM(xen_dm_op_buf_t) bufs); +extern long +do_hypfs_op( + unsigned int cmd, + XEN_GUEST_HANDLE_PARAM(void) arg1,
Do you anticipate this parameter to be used for other than path names? If not, perhaps better XEN_GUEST_HANDLE_PARAM(const_char)?
Quoted 63 lines
xen/include/xen/hypfs.h new file +60
-0@@ -0,0 +1,89 @@ +#ifndef __XEN_HYPFS_H__ +#define __XEN_HYPFS_H__ + +#include <xen/list.h> +#include <xen/string.h> +#include <public/hypfs.h> + +struct hypfs_entry_leaf ; + +struct hypfs_entry { +unsigned short type; +unsigned short encoding; +unsigned int size; +const char *name; +struct list_head list; +int (*read)(const struct hypfs_entry *entry, +XEN_GUEST_HANDLE_PARAM (void ) uaddr); +int (*write)(struct hypfs_entry_leaf *leaf, +XEN_GUEST_HANDLE_PARAM (void ) uaddr,unsigned long ulen); +}; + +struct hypfs_entry_leaf { +struct hypfs_entry e; +union { +const void *content; +void *write_ptr; + }; +}; + +struct hypfs_entry_dir { +struct hypfs_entry e; +struct list_head dirlist; +}; + +#define HYPFS_DIR_INIT (var, nam) \ + struct hypfs_entry_dir var = { \ + .e.type = XEN_HYPFS_TYPE_DIR, \ + .e.encoding = XEN_HYPFS_ENC_PLAIN, \ + .e.name = nam, \ + .e.size = 0, \ + .e.list = LIST_HEAD_INIT(var.e.list), \ + .e.read = hypfs_read_dir, \ + .dirlist = LIST_HEAD_INIT(var.dirlist), \ + } + +/* Content and size need to be set via hypfs_string_set(). */ +#define HYPFS_STRING_INIT (var, nam) \ + struct hypfs_entry_leaf var = { \ + .e.type = XEN_HYPFS_TYPE_STRING, \ + .e.encoding = XEN_HYPFS_ENC_PLAIN, \ + .e.name = nam, \ + .e.read = hypfs_read_leaf, \ + } + +static inline void hypfs_string_set (struct hypfs_entry_leaf *leaf, +const char *str) +{ + leaf->content = str; + leaf->e.size =strlen (str) +1 ; +}
xen/include/xen/hypfs.h line 70
Quoted 10 lines
xen/include/xen/hypfs.h new file +10
-0+ +#define HYPFS_UINT_INIT (var, nam, uint) \ + struct hypfs_entry_leaf var = { \ + .e.type = XEN_HYPFS_TYPE_UINT, \ + .e.encoding = XEN_HYPFS_ENC_PLAIN, \ + .e.name = nam, \ + .e.size = sizeof(uint), \ + .e.read = hypfs_read_leaf, \ + .content = &uint, \ + }
So you've got such helper macros for dir, string, and uint. Why not e.g. int and bool?
xen/include/xen/hypfs.h line 73
Quoted 3 lines
xen/include/xen/hypfs.h new file +3
-0+ + +extern struct hypfs_entry_dir hypfs_root;
No double blank lines please.
xen/include/xen/hypfs.h line 75
Quoted 1 line
xen/include/xen/hypfs.h new file +1
-0+struct hypfs_entry *hypfs_get_entry (const char *path);
Does the only caller really need a non-const return type? Even hypfs_write() doesn't look to modify what its leaf parameter points at.
And is there indeed an expectation for this to be used from outside of the source file it's defined in?
Jan
_______________________________________________
Xen-devel mailing list
[email protected]
https://lists.xenproject.org/mailman/listinfo/xen-devel
Re: [Xen-devel] [PATCH v3 4/9] xen: add basic hypervisor filesystem support
Quoted 14 lines · Jan Beulich
Quoted 4 lines · Juergen Gross
--- xen/arch/arm/traps.c | 1 + xen/arch/x86/hvm/hypercall.c | 1 + xen/arch/x86/hypercall.c | 1 +Quoted 8 lines
xen/arch/x86/pv/hypercall.c | 1 + xen/common/Makefile | 1 + xen/common/hypfs.c | 365 +++++++++++++++++++++++++++++++++++++++++++ xen/include/public/hypfs.h | 124 +++++++++++++++ xen/include/public/xen.h | 1 + xen/include/xen/hypercall.h | 8 + xen/include/xen/hypfs.h | 89 +++++++++++ 10 files changed, 592 insertions(+)
Quoted 4 lines
Even if it's just two structures that you have in the public header, your assertion of the interface being guest bitness agnostic should be accompanied by actual proof, i.e. addition to xen/include/xlat.lst.
Okay.
Quoted 30 lines
Quoted 30 lines
xen/common/hypfs.c new file +29
-0+static int add_entry (struct hypfs_entry_dir *parent,struct hypfs_entry *new) +{ +int ret = -ENOENT ; +struct hypfs_entry *e; + +write_lock (&hypfs_lock); + + list_for_each_entry ( e, &parent->dirlist, list ) + { +int cmp =strcmp (e->name, new->name); + +if ( cmp >0 ) + { + ret =0 ; +list_add_tail (&new->list, &e->list); +break ; + } +if ( cmp ==0 ) + { + ret = -EEXIST ; +break ; + } + } + +if ( ret == -ENOENT ) + { + ret =0 ; +list_add_tail (&new->list, &parent->dirlist); + }
Quoted 8 lines
xen/common/hypfs.c line 58
Quoted 6 lines
xen/common/hypfs.c new file +6
-0+ +if ( !ret ) + { +unsigned int sz =strlen (new->name) +1 ; + + parent->e.size +=DIRENTRY_SIZE (sz);Would
DIRENTRY_SIZE()perhaps better include the "+ 1"?
Good idea.
Quoted 15 lines
xen/common/hypfs.c line 75
Quoted 11 lines
xen/common/hypfs.c new file +10
-0+int hypfs_add_entry (struct hypfs_entry_dir *parent, +struct hypfs_entry *entry,bool nofault) +{ +int ret; + + ret =add_entry (parent, entry); +BUG_ON (nofault && ret); + +return ret; +}While this and its two siblings have no caller, the one above doesn't even have a declaration in the header file. What is this intended to be used for (from external callers)?
Oh, this seems to be a leftover from development phase. I'll remove it.
Quoted 2 lines
I also think the "nofault" aspect could do with discussing in the commit message - it seems quite odd to me.
I realized I was repeating the same code pattern multiple times, so I decidied to have a common function for it. Will add something to the commit message.
Quoted 6 lines
xen/common/hypfs.c line 103
Quoted 3 lines
xen/common/hypfs.c new file +2
-0+static int hypfs_get_path_user (char *buf,XEN_GUEST_HANDLE_PARAM (void ) uaddr, +unsigned long len)For consistency with naming elsewhere as well as uaddr here - ulen?
Okay.
Quoted 16 lines
xen/common/hypfs.c line 128
Quoted 13 lines
xen/common/hypfs.c new file +12
-0+static struct hypfs_entry *hypfs_get_entry_rel (struct hypfs_entry_dir *dir, +const char *path) +{ +const char *end; +struct hypfs_entry *entry; +unsigned int name_len; + +if ( !*path ) +return &dir->e; + +if ( dir->e.type !=XEN_HYPFS_TYPE_DIR ) +return NULL ;Don't these two need switching around, to make sure /a/b/c/ doesn't match a non-dir /a/b/c ?
Oh, indeed.
Quoted 12 lines
xen/common/hypfs.c line 138
Quoted 10 lines
xen/common/hypfs.c new file +9
-0+ end =strchr (path,'/' ); +if ( !end ) + end =strchr (path,'\0' ); + name_len = end - path; + + list_for_each_entry ( entry, &dir->dirlist, list ) + { +int cmp =strncmp (path, entry->name, name_len); + struct hypfs_entry_dir *d = container_of(entry,A hard tab slipped in here.
Sorry for that.
Quoted 19 lines
Quoted 19 lines
xen/common/hypfs.c new file +18
-0+ struct hypfs_entry_dir, e); + +if ( cmp <0 ) +return NULL ; +if ( !cmp &&strlen (entry->name) == name_len ) +return *end ?hypfs_get_entry_rel (d, end +1 ) : entry; + } + +return NULL ; +} + +struct hypfs_entry *hypfs_get_entry (const char *path) +{ +if ( path[0 ] !='/' ) +return NULL ; + +return hypfs_get_entry_rel (&hypfs_root, path +1 ); +}
Quoted 8 lines
xen/common/hypfs.c line 162
Quoted 6 lines
xen/common/hypfs.c new file +6
-0+ +int hypfs_read_dir (const struct hypfs_entry *entry, +XEN_GUEST_HANDLE_PARAM (void ) uaddr) +{ +const struct hypfs_entry_dir *d; +struct hypfs_entry *e;const?
Yes.
Quoted 17 lines
Quoted 17 lines
xen/common/hypfs.c new file +16
-0+static int hypfs_read (const struct hypfs_entry *entry, +XEN_GUEST_HANDLE_PARAM (void ) uaddr,unsigned long ulen) +{ +struct xen_hypfs_direntry e; +long ret = -EINVAL ; + +if ( ulen <sizeof (e) ) + goto out; + + e.flags = entry->write ?XEN_HYPFS_WRITEABLE :0 ; + e.type = entry->type; + e.encoding = entry->encoding; + e.content_len = entry->size; + + ret = -EFAULT ; +if (copy_to_guest (uaddr, &e,1 ) )
Quoted 8 lines
xen/common/hypfs.c line 224
Quoted 5 lines
xen/common/hypfs.c new file +5
-0+ goto out; + + ret =0 ; +if ( ulen < entry->size +sizeof (e) ) + goto out;So you return "success" even if the operation didn't complete successfully. This isn't very nice, plus ...
The direntry contains the needed size. The caller should know the size he passed to Xen.
Quoted 7 lines
xen/common/hypfs.c line 228
Quoted 4 lines
xen/common/hypfs.c new file +3
-0+guest_handle_add_offset (uaddr,sizeof (e)); + + ret = entry->read (entry, uaddr);... how is the caller to know whether direntry was at least copied if this then fails?
Is this really important? Normally -EFAULT should just not happen. In case it does I don't think the caller can make real use of the direntry.
Quoted 11 lines
xen/common/hypfs.c line 241
Quoted 9 lines
xen/common/hypfs.c new file +8
-0+int hypfs_write_leaf (struct hypfs_entry_leaf *leaf, +XEN_GUEST_HANDLE_PARAM (void ) uaddr,unsigned long ulen) +{ +char *buf; +int ret; + +if ( ulen > leaf->e.size ) + ulen = leaf->e.size;Silent truncation?
Hmm, true. I'll modify it to return an error in that case.
Quoted 3 lines
Quoted 1 line
+ buf =
xzalloc_array(char, ulen);Why the z variant?
Okay, will use xmalloc().
Quoted 15 lines
xen/common/hypfs.c line 253
Quoted 11 lines
xen/common/hypfs.c new file +10
-0+if ( !buf ) +return -ENOMEM ; + + ret = -EFAULT ; +if (copy_from_guest (buf, uaddr, ulen) ) + goto out; + + ret =0 ; +if ( leaf->e.type ==XEN_HYPFS_TYPE_STRING ) + buf[leaf->e.size -1 ] =0 ;And possible further silent truncation? And if the incoming buffer has no nul byte at ulen-1, you'll then ...
Quoted 4 lines
Quoted 1 line
+
memcpy(leaf->write_ptr, buf, ulen);... produce a strange concatenation of new and tail of old contents?
Will rework by returning an error if the size is too large or there is no terminating nul byte.
Quoted 15 lines
Anyway, this and ...
xen/common/hypfs.c line 263
Quoted 8 lines
+ out:
xen/common/hypfs.c new file +7
-0+xfree (buf); +return ret; +} + +int hypfs_write_bool (struct hypfs_entry_leaf *leaf, +XEN_GUEST_HANDLE_PARAM (void ) uaddr,unsigned long ulen) +{... this function aren't very helpful to review without there being a caller. Could these be introduced at the time a first caller appears?
Of course. Question is where to stop. I wanted to have the basic hypfs support in one patch. Are you fine with just those two functions being moved to the runtime parameter patch?
Quoted 12 lines
Quoted 12 lines
xen/common/hypfs.c new file +11
-0+union { +char buf[8 ]; +uint8_t u8; +uint16_t u16; +uint32_t u32; +uint64_t u64; + } u; + +ASSERT (leaf->e.type ==XEN_HYPFS_TYPE_UINT && leaf->e.size <=8 ); + +if ( ulen != leaf->e.size )
Quoted 8 lines
Quoted 1 line
+ return
-EDOM;Is this restriction really necessary? Setting e.g. a 4-byte field from 1-byte input is no problem at all. This being for booleans I anyway wonder why input might be helpful to have larger than a single byte. But maybe all of this is again a result of not seeing what a user of the function would look like.
I wanted to have as little functionality as possible in the hypervisor. It is no problem for the library to pass a properly sized buffer.
Allowing larger variables for booleans is just a consequence of the hypervisor parameters allowing that.
Quoted 39 lines
Quoted 39 lines
xen/common/hypfs.c new file +38
-0+long do_hypfs_op (unsigned int cmd, +XEN_GUEST_HANDLE_PARAM (void ) arg1,unsigned long arg2, +XEN_GUEST_HANDLE_PARAM (void ) arg3,unsigned long arg4) +{ +int ret; +struct hypfs_entry *entry; +static char path[XEN_HYPFS_MAX_PATHLEN ]; + +if ( !is_control_domain (current->domain) && + !is_hardware_domain (current->domain) ) +return -EPERM ; + +if ( cmd == XEN_HYPFS_OP_get_version ) +return XEN_HYPFS_VERSION ; + +if ( cmd == XEN_HYPFS_OP_write_contents ) +write_lock (&hypfs_lock); +else +read_lock (&hypfs_lock); + + ret =hypfs_get_path_user (path, arg1, arg2); +if ( ret ) + goto out; + + entry =hypfs_get_entry (path); +if ( !entry ) + { + ret = -ENOENT ; + goto out; + } + +switch ( cmd ) + { +case XEN_HYPFS_OP_read_contents: + ret =hypfs_read (entry, arg3, arg4); +break ; + +case XEN_HYPFS_OP_write_contents:
Quoted 8 lines
xen/common/hypfs.c line 354
Quoted 5 lines
xen/common/hypfs.c new file +5
-0+ ret =hypfs_write (entry, arg3, arg4); +break ; + +default : + ret = -ENOSYS ;EINVAL or EOPNOTSUPP please. ENOSYS should be used only for not implemented top level functions.
Okay, will use EOPNOTSUPP.
Quoted 36 lines
Quoted 36 lines
xen/include/public/hypfs.h new file +32
-0@@ -0,0 +1,124 @@ +/****************************************************************************** +* Xen Hypervisor Filesystem +* +* Copyright (c) 2019, SUSE Software Solutions Germany GmbH +* +* Permission is hereby granted, free of charge, to any person obtaining a copy +* of this software and associated documentation files (the "Software"), to +* deal in the Software without restriction, including without limitation the +* rights to use, copy, modify, merge, publish, distribute, sublicense, and/or +* sell copies of the Software, and to permit persons to whom the Software is +* furnished to do so, subject to the following conditions: +* +* The above copyright notice and this permission notice shall be included in +* all copies or substantial portions of the Software. +* +* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR +* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, +* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE +* AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER +* LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING +* FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER +* DEALINGS IN THE SOFTWARE. +* +*/ + +#ifndef __XEN_PUBLIC_HYPFS_H__ +#define __XEN_PUBLIC_HYPFS_H__ + +#include "xen.h" + +/* +* Definitions for the __HYPERVISOR_hypfs_op hypercall.
Quoted 8 lines
xen/include/public/hypfs.h line 36
Quoted 4 lines
xen/include/public/hypfs.h new file +4
-0+*/ + +/* Highest version number of the hypfs interface currently defined. */ +#define XEN_HYPFS_VERSION 1For this and the accompanying
XEN_HYPFS_OP_get_version, at least the doc added by patch 3 could actually do with mentioning the intentions you have with this.
Okay.
Quoted 13 lines
Quoted 13 lines
xen/include/public/hypfs.h new file +12
-0+/* Maximum length of a path in the filesystem. */ +#define XEN_HYPFS_MAX_PATHLEN 1024 + +struct xen_hypfs_direntry { +uint16_t flags; +#define XEN_HYPFS_WRITEABLE 0x0001 +uint8_t type; +#define XEN_HYPFS_TYPE_DIR 0x0000 +#define XEN_HYPFS_TYPE_BLOB 0x0001 +#define XEN_HYPFS_TYPE_STRING 0x0002 +#define XEN_HYPFS_TYPE_UINT 0x0003 +#define XEN_HYPFS_TYPE_INT 0x0004
Quoted 8 lines
xen/include/public/hypfs.h line 53
Quoted 4 lines
xen/include/public/hypfs.h new file +4
-0+#define XEN_HYPFS_TYPE_BOOL 0x0005 +uint8_t encoding; +#define XEN_HYPFS_ENC_PLAIN 0x0000 +#define XEN_HYPFS_ENC_GZIP 0x0001Meaning I can e.g. have a gzip-ed string or bool (or even dir)? If this is just for "blob", why have separate fields instead of e.g.
BLOB_RAWandBLOB_GZIPor some such?
gzip-ed string or blob are the primary targets.
Maybe we want to have other encoding s later (Andrew asked for that possibility when I posted the patch for retrieving the .config file contents early last year).
Quoted 3 lines
Also - why 4 digits in the constants when the fields are
uint8_t? Since these are enum-like, I even wonder if hex is warranted here.
I'm fine with switching to plain integers.
Quoted 14 lines
xen/include/public/hypfs.h line 62
Quoted 10 lines
xen/include/public/hypfs.h new file +9
-0+uint32_t content_len; +}; + +struct xen_hypfs_dirlistentry { +struct xen_hypfs_direntry e; +/* Offset in bytes to next entry (0 == this is the last entry). */ +uint16_t off_next; +char name[XEN_FLEX_ARRAY_DIM ]; +};The interaction of the last two fields may want spelling out:
I _assume_ name[] is nul-terminated, andoff_nextexists to potentially skip trailing padding?
Yes. Will add a comment.
Quoted 24 lines
Quoted 24 lines
xen/include/public/hypfs.h new file +23
-0+/* +* Hypercall operations. +*/ + +/* +* XEN_HYPFS_OP_get_version +* +* Read highest interface version supported by the hypervisor. +* +* Possible return values: +* >0: highest supported interface version +* <0: negative Xen errno value +*/ +#define XEN_HYPFS_OP_get_version 0 + +/* +* XEN_HYPFS_OP_read_contents +* +* Read contents of a filesystem entry. +* +* Returns the direntry and contents of an entry in the buffer supplied by the +* caller (struct xen_hypfs_direntry with the contents following directly +* after it).
Quoted 8 lines
xen/include/public/hypfs.h line 91
Quoted 5 lines
xen/include/public/hypfs.h new file +5
-0+* The data buffer must be at least the size of the direntry returned in order +* to have success. If the data buffer was not large enough for all the data +* no entry data is returned, but the direntry will contain the needed size +* for the returned data. +* The format of the contents is according to its entry type and encoding. What specifically this means for a dir would be nice to be spelled out.
Okay.
Quoted 28 lines
Quoted 28 lines
xen/include/public/hypfs.h new file +27
-0+* arg1: XEN_GUEST_HANDLE(path name) +* arg2: length of path name (including trailing zero byte) +* arg3: XEN_GUEST_HANDLE(data buffer written by hypervisor) +* arg4: data buffer size +* +* Possible return values: +* 0: success (at least the direntry was returned) +* <0 : negative Xen errno value +*/ +#define XEN_HYPFS_OP_read_contents 1 + +/* +* XEN_HYPFS_OP_write_contents +* +* Write contents of a filesystem entry. +* +* Writes an entry with the contents of a buffer supplied by the caller. +* The data type and encoding can't be changed. The size can be changed only +* for blobs and strings. +* +* arg1: XEN_GUEST_HANDLE(path name) +* arg2: length of path name (including trailing zero byte) +* arg3: XEN_GUEST_HANDLE(content buffer read by hypervisor) +* arg4: content buffer size +* +* Possible return values: +* 0: success
Quoted 8 lines
xen/include/public/hypfs.h line 122
Quoted 3 lines
xen/include/public/hypfs.h new file +3
-0+* <0 : negative Xen errno value +*/ +#define XEN_HYPFS_OP_write_contents 2This one indeed accesses only the actual data (contents) of the referenced entry.
XEN_HYPFS_OP_read_contentshands back also the dir entry. Should the latter then better be named justXEN_HYPFS_OP_read?
Yes, this might be better.
Quoted 7 lines
Quoted 7 lines
xen/include/xen/hypercall.h +0
-0@@ -150,6 +150,14 @@ do_dm_op( unsigned int nr_bufs, XEN_GUEST_HANDLE_PARAM(xen_dm_op_buf_t) bufs);
Quoted 7 lines
xen/include/xen/hypercall.h do_dm_op( · line 156
Quoted 4 lines
xen/include/xen/hypercall.h +4
-0+extern long +do_hypfs_op( + unsigned int cmd, + XEN_GUEST_HANDLE_PARAM(void) arg1,Do you anticipate this parameter to be used for other than path names? If not, perhaps better
XEN_GUEST_HANDLE_PARAM(const_char)?
Yes, I think this can be changed.
Quoted 69 lines
Quoted 69 lines
xen/include/xen/hypfs.h new file +65
-0@@ -0,0 +1,89 @@ +#ifndef __XEN_HYPFS_H__ +#define __XEN_HYPFS_H__ + +#include <xen/list.h> +#include <xen/string.h> +#include <public/hypfs.h> + +struct hypfs_entry_leaf ; + +struct hypfs_entry { +unsigned short type; +unsigned short encoding; +unsigned int size; +const char *name; +struct list_head list; +int (*read)(const struct hypfs_entry *entry, +XEN_GUEST_HANDLE_PARAM (void ) uaddr); +int (*write)(struct hypfs_entry_leaf *leaf, +XEN_GUEST_HANDLE_PARAM (void ) uaddr,unsigned long ulen); +}; + +struct hypfs_entry_leaf { +struct hypfs_entry e; +union { +const void *content; +void *write_ptr; + }; +}; + +struct hypfs_entry_dir { +struct hypfs_entry e; +struct list_head dirlist; +}; + +#define HYPFS_DIR_INIT (var, nam) \ + struct hypfs_entry_dir var = { \ + .e.type = XEN_HYPFS_TYPE_DIR, \ + .e.encoding = XEN_HYPFS_ENC_PLAIN, \ + .e.name = nam, \ + .e.size = 0, \ + .e.list = LIST_HEAD_INIT(var.e.list), \ + .e.read = hypfs_read_dir, \ + .dirlist = LIST_HEAD_INIT(var.dirlist), \ + } + +/* Content and size need to be set via hypfs_string_set(). */ +#define HYPFS_STRING_INIT (var, nam) \ + struct hypfs_entry_leaf var = { \ + .e.type = XEN_HYPFS_TYPE_STRING, \ + .e.encoding = XEN_HYPFS_ENC_PLAIN, \ + .e.name = nam, \ + .e.read = hypfs_read_leaf, \ + } + +static inline void hypfs_string_set (struct hypfs_entry_leaf *leaf, +const char *str) +{ + leaf->content = str; + leaf->e.size =strlen (str) +1 ; +} + +#define HYPFS_UINT_INIT (var, nam, uint) \ + struct hypfs_entry_leaf var = { \ + .e.type = XEN_HYPFS_TYPE_UINT, \ + .e.encoding = XEN_HYPFS_ENC_PLAIN, \
Quoted 8 lines
xen/include/xen/hypfs.h line 70
Quoted 5 lines
xen/include/xen/hypfs.h new file +5
-0+ .e.name = nam, \ + .e.size = sizeof(uint), \ + .e.read = hypfs_read_leaf, \ + .content = &uint, \ + }So you've got such helper macros for dir, string, and uint. Why not e.g. int and bool?
There are no users in my series yet.
Quoted 6 lines
xen/include/xen/hypfs.h line 73
Quoted 4 lines
xen/include/xen/hypfs.h new file +3
-0+ + +extern struct hypfs_entry_dir hypfs_root;No double blank lines please.
Oh, sorry.
Quoted 6 lines
xen/include/xen/hypfs.h line 75
Quoted 2 lines
xen/include/xen/hypfs.h new file +1
-0+struct hypfs_entry *hypfs_get_entry (const char *path);Does the only caller really need a non-const return type? Even
hypfs_write()doesn't look to modify what its leaf parameter points at.
This might change when support for dynamically allocated strings or blobs is added (I have no plans to do this right now, but its easy to think about the need).
Quoted 3 lines
And is there indeed an expectation for this to be used from outside of the source file it's defined in?
Yes. As soon as support for e.g. per-domain or per-cpupool nodes is added this will be needed.
Juergen
_______________________________________________
Xen-devel mailing list
[email protected]
https://lists.xenproject.org/mailman/listinfo/xen-devel
Re: [Xen-devel] [PATCH v3 4/9] xen: add basic hypervisor filesystem support
Quoted 35 lines · Jürgen Groß
Quoted 17 lines · Jan Beulich
Quoted 16 lines · Juergen Gross
xen/common/hypfs.c new file +16
-0+static int hypfs_read (const struct hypfs_entry *entry, +XEN_GUEST_HANDLE_PARAM (void ) uaddr,unsigned long ulen) +{ +struct xen_hypfs_direntry e; +long ret = -EINVAL ; + +if ( ulen <sizeof (e) ) + goto out; + + e.flags = entry->write ?XEN_HYPFS_WRITEABLE :0 ; + e.type = entry->type; + e.encoding = entry->encoding; + e.content_len = entry->size; + + ret = -EFAULT ; +if (copy_to_guest (uaddr, &e,1 ) )Quoted 8 lines
xen/common/hypfs.c line 224
Quoted 5 lines
xen/common/hypfs.c new file +5
-0+ goto out; + + ret =0 ; +if ( ulen < entry->size +sizeof (e) ) + goto out;So you return "success" even if the operation didn't complete successfully. This isn't very nice, plus ...
The direntry contains the needed size. The caller should know the size he passed to Xen.
Quoted 5 lines
Quoted 4 lines
xen/common/hypfs.c new file +3
-0+guest_handle_add_offset (uaddr,sizeof (e)); + + ret = entry->read (entry, uaddr);
Quoted 5 lines
Quoted 2 lines
... how is the caller to know whether direntry was at least copied if this then fails?
Is this really important? Normally
-EFAULTshould just not happen. In case it does I don't think the caller can make real use of the direntry.
"Important" has various possible meanings. The success/failure indication to the caller should at least be rational. "If the data buffer was not large enough for all the data no entry data is returned, but the direntry will contain the needed size for the returned data" is fine to be stated in the public header, but I think this wants to be -ENOBUFS then, not 0 (success).
Quoted 15 lines
Quoted 14 lines
Anyway, this and ...
xen/common/hypfs.c line 263
Quoted 8 lines
+ out:
xen/common/hypfs.c new file +7
-0+xfree (buf); +return ret; +} + +int hypfs_write_bool (struct hypfs_entry_leaf *leaf, +XEN_GUEST_HANDLE_PARAM (void ) uaddr,unsigned long ulen) +{... this function aren't very helpful to review without there being a caller. Could these be introduced at the time a first caller appears?
Quoted 3 lines
Of course. Question is where to stop. I wanted to have the basic hypfs support in one patch. Are you fine with just those two functions being moved to the runtime parameter patch?
Let me put it this way: For anything the patch adds but there's no usage at all (i.e. not even in a macro, where at least the usage intentions get sufficiently clarified), the description should cover for this lack of sufficient context. Therefore I'd also be fine with the two functions remaining here, as long as readers (reviewers) can understand the intentions. It might still be _easier_ for everyone to have them in a later patch. But then the same still goes for other functions that have no users here. (The helper macros HYPFS_*_INIT(), otoh, are clear enough the way they are imo, and hence are fine to remain, plus they serve as usage explanation for hypfs_read_{leaf,dir}(), which as it looks would otherwise too be orphaned.)
Quoted 20 lines
Quoted 11 lines
Quoted 11 lines
xen/common/hypfs.c new file +11
-0+union { +char buf[8 ]; +uint8_t u8; +uint16_t u16; +uint32_t u32; +uint64_t u64; + } u; + +ASSERT (leaf->e.type ==XEN_HYPFS_TYPE_UINT && leaf->e.size <=8 ); + +if ( ulen != leaf->e.size )Quoted 8 lines
Quoted 1 line
+ return
-EDOM;Is this restriction really necessary? Setting e.g. a 4-byte field from 1-byte input is no problem at all. This being for booleans I anyway wonder why input might be helpful to have larger than a single byte. But maybe all of this is again a result of not seeing what a user of the function would look like.
Quoted 5 lines
I wanted to have as little functionality as possible in the hypervisor. It is no problem for the library to pass a properly sized buffer.
Allowing larger variables for booleans is just a consequence of the hypervisor parameters allowing that.
But the caller shouldn't be concerned of the hypervisor implementation detail of what the chose width is. Over time we e.g. convert int (along with bool_t) to bool when it's used in a boolean way. This should not result in the caller needing to change, despite the width change of the variable.
Quoted 68 lines
Quoted 35 lines
Quoted 35 lines
xen/include/public/hypfs.h new file +32
-0@@ -0,0 +1,124 @@ +/****************************************************************************** +* Xen Hypervisor Filesystem +* +* Copyright (c) 2019, SUSE Software Solutions Germany GmbH +* +* Permission is hereby granted, free of charge, to any person obtaining a copy +* of this software and associated documentation files (the "Software"), to +* deal in the Software without restriction, including without limitation the +* rights to use, copy, modify, merge, publish, distribute, sublicense, and/or +* sell copies of the Software, and to permit persons to whom the Software is +* furnished to do so, subject to the following conditions: +* +* The above copyright notice and this permission notice shall be included in +* all copies or substantial portions of the Software. +* +* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR +* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, +* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE +* AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER +* LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING +* FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER +* DEALINGS IN THE SOFTWARE. +* +*/ + +#ifndef __XEN_PUBLIC_HYPFS_H__ +#define __XEN_PUBLIC_HYPFS_H__ + +#include "xen.h" + +/* +* Definitions for the __HYPERVISOR_hypfs_op hypercall. Quoted 8 lines
xen/include/public/hypfs.h line 36
Quoted 4 lines
xen/include/public/hypfs.h new file +4
-0+*/ + +/* Highest version number of the hypfs interface currently defined. */ +#define XEN_HYPFS_VERSION 1For this and the accompanying
XEN_HYPFS_OP_get_version, at least the doc added by patch 3 could actually do with mentioning the intentions you have with this.Okay.
Quoted 13 lines
Quoted 13 lines
xen/include/public/hypfs.h new file +12
-0+/* Maximum length of a path in the filesystem. */ +#define XEN_HYPFS_MAX_PATHLEN 1024 + +struct xen_hypfs_direntry { +uint16_t flags; +#define XEN_HYPFS_WRITEABLE 0x0001 +uint8_t type; +#define XEN_HYPFS_TYPE_DIR 0x0000 +#define XEN_HYPFS_TYPE_BLOB 0x0001 +#define XEN_HYPFS_TYPE_STRING 0x0002 +#define XEN_HYPFS_TYPE_UINT 0x0003 +#define XEN_HYPFS_TYPE_INT 0x0004Quoted 8 lines
xen/include/public/hypfs.h line 53
Quoted 4 lines
xen/include/public/hypfs.h new file +4
-0+#define XEN_HYPFS_TYPE_BOOL 0x0005 +uint8_t encoding; +#define XEN_HYPFS_ENC_PLAIN 0x0000 +#define XEN_HYPFS_ENC_GZIP 0x0001Meaning I can e.g. have a gzip-ed string or bool (or even dir)? If this is just for "blob", why have separate fields instead of e.g.
BLOB_RAWandBLOB_GZIPor some such?
Quoted 5 lines
gzip-ed string or blob are the primary targets.
Maybe we want to have other encoding s later (Andrew asked for that possibility when I posted the patch for retrieving the .config file contents early last year).
To me it would seem preferable if the contents of a blob identified itself as to its format. But since this leaves room for ambiguities I accept that the format needs specifying. However, to me a gzip-ed string is as good as a gzip-ed blob, and hence I still think sub-dividing "blob" is the way to go, with no separate "encoding". Otherwise at the very least a comment here would need adding to clarify what combinations are valid / to be expected by callers.
Quoted 40 lines
Quoted 31 lines
Quoted 31 lines
xen/include/public/hypfs.h new file +31
-0+#define HYPFS_DIR_INIT (var, nam) \ + struct hypfs_entry_dir var = { \ + .e.type = XEN_HYPFS_TYPE_DIR, \ + .e.encoding = XEN_HYPFS_ENC_PLAIN, \ + .e.name = nam, \ + .e.size = 0, \ + .e.list = LIST_HEAD_INIT(var.e.list), \ + .e.read = hypfs_read_dir, \ + .dirlist = LIST_HEAD_INIT(var.dirlist), \ + } + +/* Content and size need to be set via hypfs_string_set(). */ +#define HYPFS_STRING_INIT (var, nam) \ + struct hypfs_entry_leaf var = { \ + .e.type = XEN_HYPFS_TYPE_STRING, \ + .e.encoding = XEN_HYPFS_ENC_PLAIN, \ + .e.name = nam, \ + .e.read = hypfs_read_leaf, \ + } + +static inline void hypfs_string_set (struct hypfs_entry_leaf *leaf, +const char *str) +{ + leaf->content = str; + leaf->e.size =strlen (str) +1 ; +} + +#define HYPFS_UINT_INIT (var, nam, uint) \ + struct hypfs_entry_leaf var = { \ + .e.type = XEN_HYPFS_TYPE_UINT, \ + .e.encoding = XEN_HYPFS_ENC_PLAIN, \Quoted 8 lines
xen/include/public/hypfs.h line 88
Quoted 5 lines
xen/include/public/hypfs.h new file +5
-0+ .e.name = nam, \ + .e.size = sizeof(uint), \ + .e.read = hypfs_read_leaf, \ + .content = &uint, \ + }So you've got such helper macros for dir, string, and uint. Why not e.g. int and bool?
Quoted 1 line
There are no users in my series yet.
Hmm, as per above strictly speaking it is this patch which matters, not the entire series. Hence I think you either want to supply a full set of helper macros here, or introduce the ones actually needed in the patches where they get first used.
Quoted 11 lines
Quoted 5 lines
xen/include/public/hypfs.h line 89
Quoted 1 line
xen/include/public/hypfs.h new file +1
-0+struct hypfs_entry *hypfs_get_entry (const char *path);Does the only caller really need a non-const return type? Even
hypfs_write()doesn't look to modify what its leaf parameter points at.This might change when support for dynamically allocated strings or blobs is added (I have no plans to do this right now, but its easy to think about the need).
Quoted 1 line
Quoted 5 lines
Quoted 2 lines
And is there indeed an expectation for this to be used from outside of the source file it's defined in?
Yes. As soon as support for e.g. per-domain or per-cpupool nodes is added this will be needed.
Until then, make the function both static and return ptr-to-const? Such that when this changes, the need for either can actually be seen?
Jan
_______________________________________________
Xen-devel mailing list
[email protected]
https://lists.xenproject.org/mailman/listinfo/xen-devel
Re: [Xen-devel] [PATCH v3 4/9] xen: add basic hypervisor filesystem support
Quoted 42 lines · Jan Beulich
Quoted 35 lines · Jürgen Groß
Quoted 17 lines · Jan Beulich
Quoted 16 lines · Juergen Gross
xen/common/hypfs.c new file +16
-0+static int hypfs_read (const struct hypfs_entry *entry, +XEN_GUEST_HANDLE_PARAM (void ) uaddr,unsigned long ulen) +{ +struct xen_hypfs_direntry e; +long ret = -EINVAL ; + +if ( ulen <sizeof (e) ) + goto out; + + e.flags = entry->write ?XEN_HYPFS_WRITEABLE :0 ; + e.type = entry->type; + e.encoding = entry->encoding; + e.content_len = entry->size; + + ret = -EFAULT ; +if (copy_to_guest (uaddr, &e,1 ) )Quoted 8 lines
xen/common/hypfs.c line 224
Quoted 5 lines
xen/common/hypfs.c new file +5
-0+ goto out; + + ret =0 ; +if ( ulen < entry->size +sizeof (e) ) + goto out;So you return "success" even if the operation didn't complete successfully. This isn't very nice, plus ...
The direntry contains the needed size. The caller should know the size he passed to Xen.
Quoted 5 lines
Quoted 4 lines
xen/common/hypfs.c new file +3
-0+guest_handle_add_offset (uaddr,sizeof (e)); + + ret = entry->read (entry, uaddr);Quoted 5 lines
Quoted 2 lines
... how is the caller to know whether direntry was at least copied if this then fails?
Is this really important? Normally
-EFAULTshould just not happen. In case it does I don't think the caller can make real use of the direntry.
Quoted 6 lines
"Important" has various possible meanings. The success/failure indication to the caller should at least be rational. "If the data buffer was not large enough for all the data no entry data is returned, but the direntry will contain the needed size for the returned data" is fine to be stated in the public header, but I think this wants to be
-ENOBUFSthen, not 0 (success).
I would be fine with this, but this contradicts your previous demand not to enumerate the possible failure cases, which would be essential for this case.
Quoted 20 lines
Quoted 16 lines
Quoted 15 lines
Anyway, this and ...
xen/common/hypfs.c line 263
Quoted 8 lines
+ out:
xen/common/hypfs.c new file +7
-0+xfree (buf); +return ret; +} + +int hypfs_write_bool (struct hypfs_entry_leaf *leaf, +XEN_GUEST_HANDLE_PARAM (void ) uaddr,unsigned long ulen) +{... this function aren't very helpful to review without there being a caller. Could these be introduced at the time a first caller appears?
Quoted 3 lines
Of course. Question is where to stop. I wanted to have the basic hypfs support in one patch. Are you fine with just those two functions being moved to the runtime parameter patch?
Quoted 12 lines
Let me put it this way: For anything the patch adds but there's no usage at all (i.e. not even in a macro, where at least the usage intentions get sufficiently clarified), the description should cover for this lack of sufficient context. Therefore I'd also be fine with the two functions remaining here, as long as readers (reviewers) can understand the intentions. It might still be _easier_ for everyone to have them in a later patch. But then the same still goes for other functions that have no users here. (The helper macros HYPFS_*_INIT(), otoh, are clear enough the way they are imo, and hence are fine to remain, plus they serve as usage explanation for hypfs_read_{leaf,dir}(), which as it looks would otherwise too be orphaned.)
Fair enough.
Quoted 24 lines
Quoted 24 lines
Quoted 12 lines
Quoted 12 lines
xen/common/hypfs.c new file +11
-0+union { +char buf[8 ]; +uint8_t u8; +uint16_t u16; +uint32_t u32; +uint64_t u64; + } u; + +ASSERT (leaf->e.type ==XEN_HYPFS_TYPE_UINT && leaf->e.size <=8 ); + +if ( ulen != leaf->e.size )Quoted 8 lines
Quoted 1 line
+ return
-EDOM;Is this restriction really necessary? Setting e.g. a 4-byte field from 1-byte input is no problem at all. This being for booleans I anyway wonder why input might be helpful to have larger than a single byte. But maybe all of this is again a result of not seeing what a user of the function would look like.
I wanted to have as little functionality as possible in the hypervisor. It is no problem for the library to pass a properly sized buffer.
Quoted 8 lines
Quoted 2 lines
Allowing larger variables for booleans is just a consequence of the hypervisor parameters allowing that.
But the caller shouldn't be concerned of the hypervisor implementation detail of what the chose width is. Over time we e.g. convert int (along with
bool_t) to bool when it's used in a boolean way. This should not result in the caller needing to change, despite the width change of the variable.
This is basically a consequence of now passing binary values to and from the hypervisor.
The normal way of handling this (as can be seen in libxenhypfs) is to query the hypervisor for the size of the value (no matter whether its int, uint or bool) and then to do the conversion between ASCII and the binary value at the caller's side.
Quoted 75 lines
Quoted 69 lines
Quoted 36 lines
Quoted 36 lines
xen/include/public/hypfs.h new file +32
-0@@ -0,0 +1,124 @@ +/****************************************************************************** +* Xen Hypervisor Filesystem +* +* Copyright (c) 2019, SUSE Software Solutions Germany GmbH +* +* Permission is hereby granted, free of charge, to any person obtaining a copy +* of this software and associated documentation files (the "Software"), to +* deal in the Software without restriction, including without limitation the +* rights to use, copy, modify, merge, publish, distribute, sublicense, and/or +* sell copies of the Software, and to permit persons to whom the Software is +* furnished to do so, subject to the following conditions: +* +* The above copyright notice and this permission notice shall be included in +* all copies or substantial portions of the Software. +* +* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR +* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, +* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE +* AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER +* LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING +* FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER +* DEALINGS IN THE SOFTWARE. +* +*/ + +#ifndef __XEN_PUBLIC_HYPFS_H__ +#define __XEN_PUBLIC_HYPFS_H__ + +#include "xen.h" + +/* +* Definitions for the __HYPERVISOR_hypfs_op hypercall. Quoted 8 lines
xen/include/public/hypfs.h line 36
Quoted 4 lines
xen/include/public/hypfs.h new file +4
-0+*/ + +/* Highest version number of the hypfs interface currently defined. */ +#define XEN_HYPFS_VERSION 1For this and the accompanying
XEN_HYPFS_OP_get_version, at least the doc added by patch 3 could actually do with mentioning the intentions you have with this.Okay.
Quoted 13 lines
Quoted 13 lines
xen/include/public/hypfs.h new file +12
-0+/* Maximum length of a path in the filesystem. */ +#define XEN_HYPFS_MAX_PATHLEN 1024 + +struct xen_hypfs_direntry { +uint16_t flags; +#define XEN_HYPFS_WRITEABLE 0x0001 +uint8_t type; +#define XEN_HYPFS_TYPE_DIR 0x0000 +#define XEN_HYPFS_TYPE_BLOB 0x0001 +#define XEN_HYPFS_TYPE_STRING 0x0002 +#define XEN_HYPFS_TYPE_UINT 0x0003 +#define XEN_HYPFS_TYPE_INT 0x0004Quoted 8 lines
xen/include/public/hypfs.h line 53
Quoted 4 lines
xen/include/public/hypfs.h new file +4
-0+#define XEN_HYPFS_TYPE_BOOL 0x0005 +uint8_t encoding; +#define XEN_HYPFS_ENC_PLAIN 0x0000 +#define XEN_HYPFS_ENC_GZIP 0x0001Meaning I can e.g. have a gzip-ed string or bool (or even dir)? If this is just for "blob", why have separate fields instead of e.g.
BLOB_RAWandBLOB_GZIPor some such?Quoted 5 lines
gzip-ed string or blob are the primary targets.
Maybe we want to have other encoding s later (Andrew asked for that possibility when I posted the patch for retrieving the .config file contents early last year).
Quoted 8 lines
To me it would seem preferable if the contents of a blob identified itself as to its format. But since this leaves room for ambiguities I accept that the format needs specifying. However, to me a gzip-ed string is as good as a gzip-ed blob, and hence I still think sub-dividing "blob" is the way to go, with no separate "encoding". Otherwise at the very least a comment here would need adding to clarify what combinations are valid / to be expected by callers.
libxenhypfs is able to handle all possible combinations. I just don't think some of the combinations are making sense (gzip-ing a binary value of 4 bytes e.g. is nonsense).
OTOH in case we'll add large arrays of longs in the future it might be beneficial to compress them in some way. So I'd like to keep type and encoding as separate information.
Quoted 41 lines
Quoted 41 lines
Quoted 40 lines
Quoted 27 lines
xen/include/public/hypfs.h new file +26
-0+#define HYPFS_DIR_INIT (var, nam) \ + struct hypfs_entry_dir var = { \ + .e.type = XEN_HYPFS_TYPE_DIR, \ + .e.encoding = XEN_HYPFS_ENC_PLAIN, \ + .e.name = nam, \ + .e.size = 0, \ + .e.list = LIST_HEAD_INIT(var.e.list), \ + .e.read = hypfs_read_dir, \ + .dirlist = LIST_HEAD_INIT(var.dirlist), \ + } + +/* Content and size need to be set via hypfs_string_set(). */ +#define HYPFS_STRING_INIT (var, nam) \ + struct hypfs_entry_leaf var = { \ + .e.type = XEN_HYPFS_TYPE_STRING, \ + .e.encoding = XEN_HYPFS_ENC_PLAIN, \ + .e.name = nam, \ + .e.read = hypfs_read_leaf, \ + } + +static inline void hypfs_string_set (struct hypfs_entry_leaf *leaf, +const char *str) +{ + leaf->content = str; + leaf->e.size =strlen (str) +1 ; +}xen/include/public/hypfs.h line 88
Quoted 10 lines
xen/include/public/hypfs.h new file +10
-0+ +#define HYPFS_UINT_INIT (var, nam, uint) \ + struct hypfs_entry_leaf var = { \ + .e.type = XEN_HYPFS_TYPE_UINT, \ + .e.encoding = XEN_HYPFS_ENC_PLAIN, \ + .e.name = nam, \ + .e.size = sizeof(uint), \ + .e.read = hypfs_read_leaf, \ + .content = &uint, \ + }So you've got such helper macros for dir, string, and uint. Why not e.g. int and bool?
Quoted 6 lines
Quoted 1 line
There are no users in my series yet.
Hmm, as per above strictly speaking it is this patch which matters, not the entire series. Hence I think you either want to supply a full set of helper macros here, or introduce the ones actually needed in the patches where they get first used.
Okay, I'll go with the full set of helpers.
Quoted 15 lines
Quoted 15 lines
Quoted 6 lines
xen/include/public/hypfs.h line 89
Quoted 2 lines
xen/include/public/hypfs.h new file +1
-0+struct hypfs_entry *hypfs_get_entry (const char *path);Does the only caller really need a non-const return type? Even
hypfs_write()doesn't look to modify what its leaf parameter points at.This might change when support for dynamically allocated strings or blobs is added (I have no plans to do this right now, but its easy to think about the need).
Quoted 3 lines
And is there indeed an expectation for this to be used from outside of the source file it's defined in?
Quoted 6 lines
Quoted 2 lines
Yes. As soon as support for e.g. per-domain or per-cpupool nodes is added this will be needed.
Until then, make the function both static and return ptr-to-const? Such that when this changes, the need for either can actually be seen?
Fine with me.
Juergen
_______________________________________________
Xen-devel mailing list
[email protected]
https://lists.xenproject.org/mailman/listinfo/xen-devel
Re: [Xen-devel] [PATCH v3 4/9] xen: add basic hypervisor filesystem support
Quoted 50 lines · Jürgen Groß
Quoted 42 lines · Jan Beulich
Quoted 35 lines · Jürgen Groß
Quoted 17 lines · Jan Beulich
Quoted 16 lines · Juergen Gross
xen/common/hypfs.c new file +16
-0+static int hypfs_read (const struct hypfs_entry *entry, +XEN_GUEST_HANDLE_PARAM (void ) uaddr,unsigned long ulen) +{ +struct xen_hypfs_direntry e; +long ret = -EINVAL ; + +if ( ulen <sizeof (e) ) + goto out; + + e.flags = entry->write ?XEN_HYPFS_WRITEABLE :0 ; + e.type = entry->type; + e.encoding = entry->encoding; + e.content_len = entry->size; + + ret = -EFAULT ; +if (copy_to_guest (uaddr, &e,1 ) )Quoted 8 lines
xen/common/hypfs.c line 224
Quoted 5 lines
xen/common/hypfs.c new file +5
-0+ goto out; + + ret =0 ; +if ( ulen < entry->size +sizeof (e) ) + goto out;So you return "success" even if the operation didn't complete successfully. This isn't very nice, plus ...
The direntry contains the needed size. The caller should know the size he passed to Xen.
Quoted 5 lines
Quoted 4 lines
xen/common/hypfs.c new file +3
-0+guest_handle_add_offset (uaddr,sizeof (e)); + + ret = entry->read (entry, uaddr);Quoted 5 lines
Quoted 2 lines
... how is the caller to know whether direntry was at least copied if this then fails?
Is this really important? Normally
-EFAULTshould just not happen. In case it does I don't think the caller can make real use of the direntry.Quoted 6 lines
"Important" has various possible meanings. The success/failure indication to the caller should at least be rational. "If the data buffer was not large enough for all the data no entry data is returned, but the direntry will contain the needed size for the returned data" is fine to be stated in the public header, but I think this wants to be
-ENOBUFSthen, not 0 (success).
Quoted 3 lines
I would be fine with this, but this contradicts your previous demand not to enumerate the possible failure cases, which would be essential for this case.
Slightly re-writing the part of the comment I did quote would be all that's needed afaict: "If the data buffer was not large enough for all the data -ENOBUFS and no entry data is returned, but the direntry will contain the needed size for the returned data."
Quoted 32 lines
Quoted 23 lines
Quoted 23 lines
Quoted 11 lines
Quoted 11 lines
xen/common/hypfs.c new file +11
-0+union { +char buf[8 ]; +uint8_t u8; +uint16_t u16; +uint32_t u32; +uint64_t u64; + } u; + +ASSERT (leaf->e.type ==XEN_HYPFS_TYPE_UINT && leaf->e.size <=8 ); + +if ( ulen != leaf->e.size )Quoted 8 lines
Quoted 1 line
+ return
-EDOM;Is this restriction really necessary? Setting e.g. a 4-byte field from 1-byte input is no problem at all. This being for booleans I anyway wonder why input might be helpful to have larger than a single byte. But maybe all of this is again a result of not seeing what a user of the function would look like.
I wanted to have as little functionality as possible in the hypervisor. It is no problem for the library to pass a properly sized buffer.
Quoted 8 lines
Quoted 2 lines
Allowing larger variables for booleans is just a consequence of the hypervisor parameters allowing that.
But the caller shouldn't be concerned of the hypervisor implementation detail of what the chose width is. Over time we e.g. convert int (along with
bool_t) to bool when it's used in a boolean way. This should not result in the caller needing to change, despite the width change of the variable.
Quoted 7 lines
This is basically a consequence of now passing binary values to and from the hypervisor.
The normal way of handling this (as can be seen in libxenhypfs) is to query the hypervisor for the size of the value (no matter whether its int, uint or bool) and then to do the conversion between ASCII and the binary value at the caller's side.
I can see why this is needed for e.g. integer values, but I don't see the need for booleans.
Quoted 33 lines
Quoted 24 lines
Quoted 18 lines
Quoted 9 lines
Quoted 9 lines
xen/include/public/hypfs.h new file +9
-0+struct xen_hypfs_direntry { +uint16_t flags; +#define XEN_HYPFS_WRITEABLE 0x0001 +uint8_t type; +#define XEN_HYPFS_TYPE_DIR 0x0000 +#define XEN_HYPFS_TYPE_BLOB 0x0001 +#define XEN_HYPFS_TYPE_STRING 0x0002 +#define XEN_HYPFS_TYPE_UINT 0x0003 +#define XEN_HYPFS_TYPE_INT 0x0004Quoted 8 lines
xen/include/public/hypfs.h line 53
Quoted 4 lines
xen/include/public/hypfs.h new file +4
-0+#define XEN_HYPFS_TYPE_BOOL 0x0005 +uint8_t encoding; +#define XEN_HYPFS_ENC_PLAIN 0x0000 +#define XEN_HYPFS_ENC_GZIP 0x0001Meaning I can e.g. have a gzip-ed string or bool (or even dir)? If this is just for "blob", why have separate fields instead of e.g.
BLOB_RAWandBLOB_GZIPor some such?Quoted 5 lines
gzip-ed string or blob are the primary targets.
Maybe we want to have other encoding s later (Andrew asked for that possibility when I posted the patch for retrieving the .config file contents early last year).
Quoted 8 lines
To me it would seem preferable if the contents of a blob identified itself as to its format. But since this leaves room for ambiguities I accept that the format needs specifying. However, to me a gzip-ed string is as good as a gzip-ed blob, and hence I still think sub-dividing "blob" is the way to go, with no separate "encoding". Otherwise at the very least a comment here would need adding to clarify what combinations are valid / to be expected by callers.
Quoted 7 lines
libxenhypfs is able to handle all possible combinations. I just don't think some of the combinations are making sense (gzip-ing a binary value of 4 bytes e.g. is nonsense).
OTOH in case we'll add large arrays of longs in the future it might be beneficial to compress them in some way. So I'd like to keep type and encoding as separate information.
Okay, I'm not entirely opposed. But I'd be curious if anyone else has an opinion here.
Jan
_______________________________________________
Xen-devel mailing list
[email protected]
https://lists.xenproject.org/mailman/listinfo/xen-devel
Re: [Xen-devel] [PATCH v3 4/9] xen: add basic hypervisor filesystem support
Quoted 51 lines · Jan Beulich
Quoted 50 lines · Jürgen Groß
Quoted 48 lines · Jan Beulich
Quoted 35 lines · Jürgen Groß
Quoted 17 lines · Jan Beulich
Quoted 16 lines · Juergen Gross
xen/common/hypfs.c new file +16
-0+static int hypfs_read (const struct hypfs_entry *entry, +XEN_GUEST_HANDLE_PARAM (void ) uaddr,unsigned long ulen) +{ +struct xen_hypfs_direntry e; +long ret = -EINVAL ; + +if ( ulen <sizeof (e) ) + goto out; + + e.flags = entry->write ?XEN_HYPFS_WRITEABLE :0 ; + e.type = entry->type; + e.encoding = entry->encoding; + e.content_len = entry->size; + + ret = -EFAULT ; +if (copy_to_guest (uaddr, &e,1 ) )Quoted 8 lines
xen/common/hypfs.c line 224
Quoted 5 lines
xen/common/hypfs.c new file +5
-0+ goto out; + + ret =0 ; +if ( ulen < entry->size +sizeof (e) ) + goto out;So you return "success" even if the operation didn't complete successfully. This isn't very nice, plus ...
The direntry contains the needed size. The caller should know the size he passed to Xen.
Quoted 5 lines
Quoted 4 lines
xen/common/hypfs.c new file +3
-0+guest_handle_add_offset (uaddr,sizeof (e)); + + ret = entry->read (entry, uaddr);Quoted 5 lines
Quoted 2 lines
... how is the caller to know whether direntry was at least copied if this then fails?
Is this really important? Normally
-EFAULTshould just not happen. In case it does I don't think the caller can make real use of the direntry."Important" has various possible meanings. The success/failure indication to the caller should at least be rational. "If the data buffer was not large enough for all the data no entry data is returned, but the direntry will contain the needed size for the returned data" is fine to be stated in the public header, but I think this wants to be
-ENOBUFSthen, not 0 (success).
Quoted 8 lines
Quoted 3 lines
I would be fine with this, but this contradicts your previous demand not to enumerate the possible failure cases, which would be essential for this case.
Slightly re-writing the part of the comment I did quote would be all that's needed afaict: "If the data buffer was not large enough for all the data
-ENOBUFSand no entry data is returned, but the direntry will contain the needed size for the returned data."
Okay. Fine with me.
Quoted 36 lines
Quoted 36 lines
Quoted 24 lines
Quoted 24 lines
Quoted 12 lines
Quoted 12 lines
xen/common/hypfs.c new file +11
-0+union { +char buf[8 ]; +uint8_t u8; +uint16_t u16; +uint32_t u32; +uint64_t u64; + } u; + +ASSERT (leaf->e.type ==XEN_HYPFS_TYPE_UINT && leaf->e.size <=8 ); + +if ( ulen != leaf->e.size )Quoted 8 lines
Quoted 1 line
+ return
-EDOM;Is this restriction really necessary? Setting e.g. a 4-byte field from 1-byte input is no problem at all. This being for booleans I anyway wonder why input might be helpful to have larger than a single byte. But maybe all of this is again a result of not seeing what a user of the function would look like.
I wanted to have as little functionality as possible in the hypervisor. It is no problem for the library to pass a properly sized buffer.
Quoted 8 lines
Quoted 2 lines
Allowing larger variables for booleans is just a consequence of the hypervisor parameters allowing that.
But the caller shouldn't be concerned of the hypervisor implementation detail of what the chose width is. Over time we e.g. convert int (along with
bool_t) to bool when it's used in a boolean way. This should not result in the caller needing to change, despite the width change of the variable.This is basically a consequence of now passing binary values to and from the hypervisor.
Quoted 7 lines
Quoted 4 lines
The normal way of handling this (as can be seen in libxenhypfs) is to query the hypervisor for the size of the value (no matter whether its int, uint or bool) and then to do the conversion between ASCII and the binary value at the caller's side.
I can see why this is needed for e.g. integer values, but I don't see the need for booleans.
At some level the conversion needs to be done. I'd rather do it for all types at the same level.
Quoted 38 lines
Quoted 38 lines
Quoted 25 lines
Quoted 19 lines
Quoted 10 lines
Quoted 10 lines
xen/include/public/hypfs.h new file +9
-0+struct xen_hypfs_direntry { +uint16_t flags; +#define XEN_HYPFS_WRITEABLE 0x0001 +uint8_t type; +#define XEN_HYPFS_TYPE_DIR 0x0000 +#define XEN_HYPFS_TYPE_BLOB 0x0001 +#define XEN_HYPFS_TYPE_STRING 0x0002 +#define XEN_HYPFS_TYPE_UINT 0x0003 +#define XEN_HYPFS_TYPE_INT 0x0004Quoted 8 lines
xen/include/public/hypfs.h line 53
Quoted 4 lines
xen/include/public/hypfs.h new file +4
-0+#define XEN_HYPFS_TYPE_BOOL 0x0005 +uint8_t encoding; +#define XEN_HYPFS_ENC_PLAIN 0x0000 +#define XEN_HYPFS_ENC_GZIP 0x0001Meaning I can e.g. have a gzip-ed string or bool (or even dir)? If this is just for "blob", why have separate fields instead of e.g.
BLOB_RAWandBLOB_GZIPor some such?Quoted 5 lines
gzip-ed string or blob are the primary targets.
Maybe we want to have other encoding s later (Andrew asked for that possibility when I posted the patch for retrieving the .config file contents early last year).
Quoted 8 lines
To me it would seem preferable if the contents of a blob identified itself as to its format. But since this leaves room for ambiguities I accept that the format needs specifying. However, to me a gzip-ed string is as good as a gzip-ed blob, and hence I still think sub-dividing "blob" is the way to go, with no separate "encoding". Otherwise at the very least a comment here would need adding to clarify what combinations are valid / to be expected by callers.
libxenhypfs is able to handle all possible combinations. I just don't think some of the combinations are making sense (gzip-ing a binary value of 4 bytes e.g. is nonsense).
Quoted 6 lines
Quoted 3 lines
OTOH in case we'll add large arrays of longs in the future it might be beneficial to compress them in some way. So I'd like to keep type and encoding as separate information.
Okay, I'm not entirely opposed. But I'd be curious if anyone else has an opinion here.
I think content type and transport encoding should not be mixed up. They are orthogonal to each other and so they should be handled.
Juergen
_______________________________________________
Xen-devel mailing list
[email protected]
https://lists.xenproject.org/mailman/listinfo/xen-devel
Re: [Xen-devel] [PATCH v3 4/9] xen: add basic hypervisor filesystem support
Quoted 47 lines · Jürgen Groß
Quoted 46 lines · Jan Beulich
Quoted 44 lines · Jürgen Groß
Quoted 27 lines · Jan Beulich
Quoted 20 lines · Jürgen Groß
Quoted 10 lines · Jan Beulich
Quoted 9 lines · Juergen Gross
(quoted diff) +9
-0+struct xen_hypfs_direntry { + uint16_t flags; +#define XEN_HYPFS_WRITEABLE 0x0001 + uint8_t type; +#define XEN_HYPFS_TYPE_DIR 0x0000 +#define XEN_HYPFS_TYPE_BLOB 0x0001 +#define XEN_HYPFS_TYPE_STRING 0x0002 +#define XEN_HYPFS_TYPE_UINT 0x0003 +#define XEN_HYPFS_TYPE_INT 0x0004Quoted 8 lines
Review
Quoted 4 lines
(quoted diff) +4
-0+#define XEN_HYPFS_TYPE_BOOL 0x0005 + uint8_t encoding; +#define XEN_HYPFS_ENC_PLAIN 0x0000 +#define XEN_HYPFS_ENC_GZIP 0x0001Meaning I can e.g. have a gzip-ed string or bool (or even dir)? If this is just for "blob", why have separate fields instead of e.g.
BLOB_RAWandBLOB_GZIPor some such?Quoted 5 lines
gzip-ed string or blob are the primary targets.
Maybe we want to have other encoding s later (Andrew asked for that possibility when I posted the patch for retrieving the .config file contents early last year).
Quoted 8 lines
To me it would seem preferable if the contents of a blob identified itself as to its format. But since this leaves room for ambiguities I accept that the format needs specifying. However, to me a gzip-ed string is as good as a gzip-ed blob, and hence I still think sub-dividing "blob" is the way to go, with no separate "encoding". Otherwise at the very least a comment here would need adding to clarify what combinations are valid / to be expected by callers.
libxenhypfs is able to handle all possible combinations. I just don't think some of the combinations are making sense (gzip-ing a binary value of 4 bytes e.g. is nonsense).
OTOH in case we'll add large arrays of longs in the future it might be beneficial to compress them in some way. So I'd like to keep type and encoding as separate information.
Quoted 5 lines
Quoted 2 lines
Okay, I'm not entirely opposed. But I'd be curious if anyone else has an opinion here.
I think content type and transport encoding should not be mixed up. They are orthogonal to each other and so they should be handled.
In principle I agree, but "blob" really covers anything or nothing at all. Yes, if strings are meant to be possible to be gzip-ed, then there is value in the separation. I'm not fully convinced though that such compressed strings (Are you thinking about .config here?) shouldn't simply be "blob" then, too.
Jan
_______________________________________________
Xen-devel mailing list
[email protected]
https://lists.xenproject.org/mailman/listinfo/xen-devel
Re: [Xen-devel] [PATCH v3 4/9] xen: add basic hypervisor filesystem support
Quoted 51 lines · Jan Beulich
Quoted 50 lines · Jürgen Groß
Quoted 48 lines · Jan Beulich
Quoted 37 lines · Jürgen Groß
Quoted 27 lines · Jan Beulich
Quoted 20 lines · Jürgen Groß
Quoted 10 lines · Jan Beulich
Quoted 9 lines · Juergen Gross
(quoted diff) +9
-0+struct xen_hypfs_direntry { + uint16_t flags; +#define XEN_HYPFS_WRITEABLE 0x0001 + uint8_t type; +#define XEN_HYPFS_TYPE_DIR 0x0000 +#define XEN_HYPFS_TYPE_BLOB 0x0001 +#define XEN_HYPFS_TYPE_STRING 0x0002 +#define XEN_HYPFS_TYPE_UINT 0x0003 +#define XEN_HYPFS_TYPE_INT 0x0004Quoted 8 lines
Review
Quoted 4 lines
(quoted diff) +4
-0+#define XEN_HYPFS_TYPE_BOOL 0x0005 + uint8_t encoding; +#define XEN_HYPFS_ENC_PLAIN 0x0000 +#define XEN_HYPFS_ENC_GZIP 0x0001Meaning I can e.g. have a gzip-ed string or bool (or even dir)? If this is just for "blob", why have separate fields instead of e.g.
BLOB_RAWandBLOB_GZIPor some such?Quoted 5 lines
gzip-ed string or blob are the primary targets.
Maybe we want to have other encoding s later (Andrew asked for that possibility when I posted the patch for retrieving the .config file contents early last year).
Quoted 8 lines
To me it would seem preferable if the contents of a blob identified itself as to its format. But since this leaves room for ambiguities I accept that the format needs specifying. However, to me a gzip-ed string is as good as a gzip-ed blob, and hence I still think sub-dividing "blob" is the way to go, with no separate "encoding". Otherwise at the very least a comment here would need adding to clarify what combinations are valid / to be expected by callers.
Quoted 7 lines
libxenhypfs is able to handle all possible combinations. I just don't think some of the combinations are making sense (gzip-ing a binary value of 4 bytes e.g. is nonsense).
OTOH in case we'll add large arrays of longs in the future it might be beneficial to compress them in some way. So I'd like to keep type and encoding as separate information.
Okay, I'm not entirely opposed. But I'd be curious if anyone else has an opinion here.
Quoted 8 lines
Quoted 2 lines
I think content type and transport encoding should not be mixed up. They are orthogonal to each other and so they should be handled.
In principle I agree, but "blob" really covers anything or nothing at all. Yes, if strings are meant to be possible to be gzip-ed, then there is value in the separation. I'm not fully convinced though that such compressed strings (Are you thinking about .config here?) shouldn't simply be "blob" then, too.
With a library on top of the hypercall it is easy to hide the encoding from the standard user. So even with .config being held in gzip-ed format in the hypervisor the xenhypfs tool will still just print the textual form of it when reading the associated node. This is different from sysfs or procfs of the Linux kernel, where the raw data is presented at the primary user interface. Additionally this enables us to avoid having to specify the compression format as a stable ABI. We could at any time switch to uncompressed format without problem.
Juergen
_______________________________________________
Xen-devel mailing list
[email protected]
https://lists.xenproject.org/mailman/listinfo/xen-devel
PATCH v3 5/9 libs: add libxenhypfs
In reply to Juergen GrossAdd the new library libxenhypfs for access to the hypervisor filesystem.
---
V1:
- rename to libxenhypfs
- add xenhypfs_write()
V3:
- major rework due to new hypervisor interface
- add decompression capability
--- .gitignore | 2 + tools/Rules.mk | 6 + tools/libs/Makefile | 1 + tools/libs/hypfs/Makefile | 16 ++ tools/libs/hypfs/core.c | 535 ++++++++++++++++++++++++++++++++++++ tools/libs/hypfs/include/xenhypfs.h | 75 +++++ tools/libs/hypfs/libxenhypfs.map | 10 + tools/libs/hypfs/xenhypfs.pc.in | 10 + 8 files changed, 655 insertions(+) create mode 100644 tools/libs/hypfs/Makefile create mode 100644 tools/libs/hypfs/core.c create mode 100644 tools/libs/hypfs/include/xenhypfs.h create mode 100644 tools/libs/hypfs/libxenhypfs.map create mode 100644 tools/libs/hypfs/xenhypfs.pc.in
.gitignore +2 -0
@@ -109,6 +109,8 @@ tools/libs/evtchn/headers.chk tools/libs/evtchn/xenevtchn.pc tools/libs/gnttab/headers.chk tools/libs/gnttab/xengnttab.pc +tools/libs/hypfs/headers.chk +tools/libs/hypfs/xenhypfs.pc tools/libs/call/headers.chk tools/libs/call/xencall.pc tools/libs/foreignmemory/headers.chk
tools/Rules.mk +6 -0
@@ -19,6 +19,7 @@ XEN_LIBXENGNTTAB = $(XEN_ROOT)/tools/libs/gnttab XEN_LIBXENCALL = $(XEN_ROOT)/tools/libs/call XEN_LIBXENFOREIGNMEMORY = $(XEN_ROOT)/tools/libs/foreignmemory XEN_LIBXENDEVICEMODEL = $(XEN_ROOT)/tools/libs/devicemodel +XEN_LIBXENHYPFS = $(XEN_ROOT)/tools/libs/hypfs XEN_LIBXC = $(XEN_ROOT)/tools/libxc XEN_XENLIGHT = $(XEN_ROOT)/tools/libxl # Currently libxlutil lives in the same directory as libxenlight
@@ -134,6 +135,11 @@ SHDEPS_libxendevicemodel = $(SHLIB_libxentoollog) $(SHLIB_libxentoolcore) $(SHLI LDLIBS_libxendevicemodel = $(SHDEPS_libxendevicemodel) $(XEN_LIBXENDEVICEMODEL)/libxendevicemodel$(libextension) SHLIB_libxendevicemodel = $(SHDEPS_libxendevicemodel) -Wl,-rpath-link=$(XEN_LIBXENDEVICEMODEL) +CFLAGS_libxenhypfs = -I$(XEN_LIBXENHYPFS)/include $(CFLAGS_xeninclude) +SHDEPS_libxenhypfs = $(SHLIB_libxentoollog) $(SHLIB_libxentoolcore) $(SHLIB_xencall) +LDLIBS_libxenhypfs = $(SHDEPS_libxenhypfs) $(XEN_LIBXENHYPFS)/libxenhypfs$(libextension) +SHLIB_libxenhypfs = $(SHDEPS_libxenhypfs) -Wl,-rpath-link=$(XEN_LIBXENHYPFS) + # code which compiles against libxenctrl get __XEN_TOOLS__ and # therefore sees the unstable hypercall interfaces. CFLAGS_libxenctrl = -I$(XEN_LIBXC)/include $(CFLAGS_libxentoollog) $(CFLAGS_libxenforeignmemory) $(CFLAGS_libxendevicemodel) $(CFLAGS_xeninclude) -D__XEN_TOOLS__
tools/libs/Makefile +1 -0
@@ -9,6 +9,7 @@ SUBDIRS-y += gnttab SUBDIRS-y += call SUBDIRS-y += foreignmemory SUBDIRS-y += devicemodel +SUBDIRS-y += hypfs ifeq ($(CONFIG_RUMP),y) SUBDIRS-y := toolcore
tools/libs/hypfs/Makefile new file +16 -0
@@ -0,0 +1,16 @@ +XEN_ROOT = $(CURDIR)/../../.. +include $(XEN_ROOT)/tools/Rules.mk + +MAJOR = 1 +MINOR = 0 +LIBNAME := hypfs +USELIBS := toollog toolcore call + +APPEND_LDFLAGS += -lz + +SRCS-y += core.c + +include ../libs.mk + +$(PKG_CONFIG_LOCAL): PKG_CONFIG_INCDIR = $(XEN_LIBXENHYPFS)/include +$(PKG_CONFIG_LOCAL): PKG_CONFIG_CFLAGS_LOCAL = $(CFLAGS_xeninclude)
tools/libs/hypfs/core.c new file +535 -0
535 lines not on this page · raw message
tools/libs/hypfs/include/xenhypfs.h new file +75 -0
75 lines not on this page · raw message
tools/libs/hypfs/libxenhypfs.map new file +10 -0
@@ -0,0 +1,10 @@ +VERS_1.0 { + global: + xenhypfs_open; + xenhypfs_close; + xenhypfs_read_raw; + xenhypfs_read; + xenhypfs_readdir; + xenhypfs_write; + local: *; /* Do not expose anything by default */ +};
tools/libs/hypfs/xenhypfs.pc.in new file +10 -0
@@ -0,0 +1,10 @@ +prefix=@@prefix@@ +includedir=@@incdir@@ +libdir=@@libdir@@ + +Name: Xenhypfs +Description: The Xenhypfs library for Xen hypervisor +Version: @@version@@ +Cflags: -I${includedir} @@cflagslocal@@ +Libs: @@libsflag@@${libdir} -lxenhypfs +Requires.private: xentoolcore,xentoollog,xencall
Signature · 7 lines
2.16.4 _______________________________________________ Xen-devel mailing list [email protected] https://lists.xenproject.org/mailman/listinfo/xen-devel
Re: [Xen-devel] [PATCH v3 5/9] libs: add libxenhypfs
Quoted 3 lines · Juergen Gross
Add the new library libxenhypfs for access to the hypervisor filesystem.
Signed-off-by Juergen Gross <[email protected]>
I've only skimmed read it. The code looks sensible.
One minor comment.
Quoted 6 lines
tools/libs/hypfs/xenhypfs.pc.in new file +0
-0
tools/libs/hypfs/xenhypfs.pc.in line 10
Quoted 10 lines
tools/libs/hypfs/xenhypfs.pc.in new file +10
-0@@ -0,0 +1,10 @@ +prefix=@@prefix@@ +includedir=@@incdir@@ +libdir=@@libdir@@ + +Name: Xenhypfs +Description: The Xenhypfs library for Xen hypervisor +Version: @@version@@ +Cflags: -I${includedir} @@cflagslocal@@ +Libs: @@libsflag@@${libdir} -lxenhypfs +Requires.private: xentoolcore,xentoollog,xencall
Need to list libz here?
_______________________________________________
Xen-devel mailing list
[email protected]
https://lists.xenproject.org/mailman/listinfo/xen-devel
Re: [Xen-devel] [PATCH v3 5/9] libs: add libxenhypfs
Quoted 19 lines · Wei Liu
Quoted 3 lines · Juergen Gross
Add the new library libxenhypfs for access to the hypervisor filesystem.
Signed-off-by Juergen Gross <[email protected]>I've only skimmed read it. The code looks sensible.
One minor comment.
Quoted 10 lines
tools/libs/hypfs/xenhypfs.pc.in new file +4
-0@@ -0,0 +1,10 @@ +prefix=@@prefix@@ +includedir=@@incdir@@ +libdir=@@libdir@@ +
Quoted 8 lines
tools/libs/hypfs/xenhypfs.pc.in line 10
Quoted 6 lines
tools/libs/hypfs/xenhypfs.pc.in new file +6
-0+Name: Xenhypfs +Description: The Xenhypfs library for Xen hypervisor +Version: @@version@@ +Cflags: -I${includedir} @@cflagslocal@@ +Libs: @@libsflag@@${libdir} -lxenhypfs +Requires.private: xentoolcore,xentoollog,xencallNeed to list libz here?
Probably, yes.
Juergen
_______________________________________________
Xen-devel mailing list
[email protected]
https://lists.xenproject.org/mailman/listinfo/xen-devel
Re: [Xen-devel] [PATCH v3 5/9] libs: add libxenhypfs
Quoted 5 lines · Jürgen Groß
Quoted 3 lines · Wei Liu
Quoted 1 line · Juergen Gross
+Requires.private: xentoolcore,xentoollog,xencall
Need to list libz here?
Probably, yes.
See "rpm -qf --provides /usr/include/zlib.h" for the proper value within the "pkgconfig" namespace. It is 'zlib' instead of 'z'.
Olaf
PATCH v3 6/9 tools: add xenfs tool
In reply to Juergen GrossAdd the xenfs tool for accessing the hypervisor filesystem.
---
V1:
- rename to xenhypfs
- don't use "--" for subcommands
- add write support
V2:
- escape non-printable characters per default with cat subcommand (Ian Jackson)
- add -b option to cat subcommand (Ian Jackson)
- add man page
V3:
- adapt to new hypfs interface
--- .gitignore | 1 + docs/man/xenhypfs.1.pod | 61 ++++++++++++++++ tools/misc/Makefile | 6 ++ tools/misc/xenhypfs.c | 189 ++++++++++++++++++++++++++++++++++++++++++++++++ 4 files changed, 257 insertions(+) create mode 100644 docs/man/xenhypfs.1.pod create mode 100644 tools/misc/xenhypfs.c
.gitignore +1 -0
@@ -367,6 +367,7 @@ tools/libxl/test_timedereg tools/libxl/test_fdderegrace tools/firmware/etherboot/eb-roms.h tools/firmware/etherboot/gpxe-git-snapshot.tar.gz +tools/misc/xenhypfs tools/misc/xenwatchdogd tools/misc/xen-hvmcrash tools/misc/xen-lowmemd
docs/man/xenhypfs.1.pod new file +61 -0
61 lines not on this page · raw message
tools/misc/Makefile +6 -0
@@ -24,6 +24,7 @@ INSTALL_SBIN-$(CONFIG_X86) += xen-lowmemd INSTALL_SBIN-$(CONFIG_X86) += xen-mfndump INSTALL_SBIN-$(CONFIG_X86) += xen-ucode INSTALL_SBIN += xencov +INSTALL_SBIN += xenhypfs INSTALL_SBIN += xenlockprof INSTALL_SBIN += xenperf INSTALL_SBIN += xenpm
@@ -86,6 +87,9 @@ xenperf: xenperf.o xenpm: xenpm.o $(CC) $(LDFLAGS) -o $@ $< $(LDLIBS_libxenctrl) $(APPEND_LDFLAGS) +xenhypfs: xenhypfs.o + $(CC) $(LDFLAGS) -o $@ $< $(LDLIBS_libxenhypfs) $(APPEND_LDFLAGS) + xenlockprof: xenlockprof.o $(CC) $(LDFLAGS) -o $@ $< $(LDLIBS_libxenctrl) $(APPEND_LDFLAGS)
@@ -94,6 +98,8 @@ xen-hptool.o: CFLAGS += -I$(XEN_ROOT)/tools/libxc $(CFLAGS_libxencall) xen-hptool: xen-hptool.o $(CC) $(LDFLAGS) -o $@ $< $(LDLIBS_libxenevtchn) $(LDLIBS_libxenctrl) $(LDLIBS_libxenguest) $(LDLIBS_libxenstore) $(APPEND_LDFLAGS) +xenhypfs.o: CFLAGS += $(CFLAGS_libxenhypfs) + # xen-mfndump incorrectly uses libxc internals xen-mfndump.o: CFLAGS += -I$(XEN_ROOT)/tools/libxc $(CFLAGS_libxencall) xen-mfndump: xen-mfndump.o
tools/misc/xenhypfs.c new file +189 -0
189 lines not on this page · raw message
Signature · 7 lines
2.16.4 _______________________________________________ Xen-devel mailing list [email protected] https://lists.xenproject.org/mailman/listinfo/xen-devel
Re: [Xen-devel] [PATCH v3 6/9] tools: add xenfs tool
Quoted 3 lines · Juergen Gross
Add the xenfs tool for accessing the hypervisor filesystem.
Signed-off-by Juergen Gross <[email protected]>
_______________________________________________
Xen-devel mailing list
[email protected]
https://lists.xenproject.org/mailman/listinfo/xen-devel
PATCH v3 7/9 xen: provide version information in hypfs
In reply to Juergen GrossProvide version and compile information in /buildinfo/ node of the Xen hypervisor file system. As this information is accessible by dom0 only no additional security problem arises.
---
V3:
- new patch
--- docs/misc/hypfs-paths.pandoc | 45 ++++++++++++++++++++++++++++++++++++++++++++ xen/common/kernel.c | 45 ++++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 90 insertions(+)
docs/misc/hypfs-paths.pandoc +45 -0
@@ -93,3 +93,48 @@ A populated Xen hypervisor file system might look like the following example: #### / The root of the hypervisor file system. + +#### /buildinfo/ + +A directory containing static information generated while building the +hypervisor. + +#### /buildinfo/changeset = STRING + +Git commit of the hypervisor. + +#### /buildinfo/compileinfo/ + +A directory containing information about compilation of Xen. + +#### /buildinfo/compileinfo/compile_by = STRING + +Information who compiled the hypervisor. + +#### /buildinfo/compileinfo/compile_date = STRING + +Date of the hypervisor compilation. + +#### /buildinfo/compileinfo/compile_domain = STRING + +Information about the compile domain. + +#### /buildinfo/compileinfo/compiler = STRING + +The compiler used to build Xen. + +#### /buildinfo/version/ + +A directory containing version information of the hypervisor. + +#### /buildinfo/version/extra = STRING + +Extra version information. + +#### /buildinfo/version/major = INTEGER + +The major version of Xen. + +#### /buildinfo/version/minor = INTEGER + +The minor version of Xen.
xen/common/kernel.c +45 -0
57 lines not on this page · raw message
Signature · 7 lines
2.16.4 _______________________________________________ Xen-devel mailing list [email protected] https://lists.xenproject.org/mailman/listinfo/xen-devel
Re: [Xen-devel] [PATCH v3 7/9] xen: provide version information in hypfs
Quoted 5 lines · Juergen Gross
Provide version and compile information in /buildinfo/ node of the Xen hypervisor file system. As this information is accessible by dom0 only no additional security problem arises.
Signed-off-by Juergen Gross <[email protected]>
with on suggestion:
Quoted 8 lines
xen/common/kernel.c +4
-0@@ -373,6 +374,50 @@ void __init do_initcalls(void) (*call)(); } +static unsigned int major_version; +static unsigned int minor_version; + +static HYPFS_DIR_INIT(buildinfo, "buildinfo");
xen/common/kernel.c void __init do_initcalls(void) · line 390
Quoted 10 lines
xen/common/kernel.c +10
-0+static HYPFS_DIR_INIT(compileinfo, "compileinfo"); +static HYPFS_DIR_INIT(version, "version"); +static HYPFS_UINT_INIT(major, "major", major_version); +static HYPFS_UINT_INIT(minor, "minor", minor_version); +static HYPFS_STRING_INIT(changeset, "changeset"); +static HYPFS_STRING_INIT(compiler, "compiler"); +static HYPFS_STRING_INIT(compile_by, "compile_by"); +static HYPFS_STRING_INIT(compile_date, "compile_date"); +static HYPFS_STRING_INIT(compile_domain, "compile_domain"); +static HYPFS_STRING_INIT(extra, "extra");
Please consider making all of the above __read_mostly.
Jan
_______________________________________________
Xen-devel mailing list
[email protected]
https://lists.xenproject.org/mailman/listinfo/xen-devel
Re: [Xen-devel] [PATCH v3 7/9] xen: provide version information in hypfs
Quoted 14 lines · Jan Beulich
Quoted 5 lines · Juergen Gross
Provide version and compile information in /buildinfo/ node of the Xen hypervisor file system. As this information is accessible by dom0 only no additional security problem arises.
Signed-off-by Juergen Gross <[email protected]>Reviewed-by Jan Beulich <[email protected]>with on suggestion:
Quoted 4 lines
xen/common/kernel.c +0
-0@@ -373,6 +374,50 @@ void __init do_initcalls(void) (*call)(); }
Quoted 16 lines
Quoted 4 lines
xen/common/kernel.c +4
-0+static unsigned int major_version; +static unsigned int minor_version; + +static HYPFS_DIR_INIT(buildinfo, "buildinfo");xen/common/kernel.c void __init do_initcalls(void) · line 390
Quoted 10 lines
xen/common/kernel.c +10
-0+static HYPFS_DIR_INIT(compileinfo, "compileinfo"); +static HYPFS_DIR_INIT(version, "version"); +static HYPFS_UINT_INIT(major, "major", major_version); +static HYPFS_UINT_INIT(minor, "minor", minor_version); +static HYPFS_STRING_INIT(changeset, "changeset"); +static HYPFS_STRING_INIT(compiler, "compiler"); +static HYPFS_STRING_INIT(compile_by, "compile_by"); +static HYPFS_STRING_INIT(compile_date, "compile_date"); +static HYPFS_STRING_INIT(compile_domain, "compile_domain"); +static HYPFS_STRING_INIT(extra, "extra");Please consider making all of the above
__read_mostly.
Okay.
Juergen
_______________________________________________
Xen-devel mailing list
[email protected]
https://lists.xenproject.org/mailman/listinfo/xen-devel
PATCH v3 8/9 xen: add /buildinfo/config entry to hypervisor filesystem
In reply to Juergen GrossAdd the /buildinfo/config entry to the hypervisor filesystem. This entry contains the .config file used to build the hypervisor.
---
V3:
- store data in gzip format
- use binfile mechanism to create data file
- move code to kernel.c
--- .gitignore | 2 ++ docs/misc/hypfs-paths.pandoc | 4 ++++ xen/common/Makefile | 12 ++++++++++++ xen/common/kernel.c | 10 ++++++++++ xen/include/xen/kernel.h | 3 +++ 5 files changed, 31 insertions(+)
.gitignore +2 -0
@@ -297,6 +297,8 @@ xen/arch/*/efi/boot.c xen/arch/*/efi/compat.c xen/arch/*/efi/efi.h xen/arch/*/efi/runtime.c +xen/common/config_data.S +xen/common/config.gz xen/include/headers*.chk xen/include/asm xen/include/asm-*/asm-offsets.h
docs/misc/hypfs-paths.pandoc +4 -0
@@ -123,6 +123,10 @@ Information about the compile domain. The compiler used to build Xen. +#### /buildinfo/config = STRING + +The contents of the `xen/.config` file at the time of the hypervisor build. + #### /buildinfo/version/ A directory containing version information of the hypervisor.
xen/common/Makefile +12 -0
@@ -1,6 +1,7 @@ obj-$(CONFIG_ARGO) += argo.o obj-y += bitmap.o obj-y += bsearch.o +obj-y += config_data.o obj-$(CONFIG_CORE_PARKING) += core_parking.o obj-y += cpu.o obj-y += cpupool.o
@@ -79,3 +80,14 @@ subdir-$(CONFIG_UBSAN) += ubsan subdir-$(CONFIG_NEEDS_LIBELF) += libelf subdir-$(CONFIG_HAS_DEVICE_TREE) += libfdt + +config.gz: ../.config + gzip -c $< >$@ + +config_data.o: config.gz + +config_data.S: $(XEN_ROOT)/xen/tools/binfile + $(XEN_ROOT)/xen/tools/binfile $@ config.gz xen_config_data + +clean:: + rm config_data.S config.gz 2>/dev/null || true
xen/common/kernel.c +10 -0
@@ -388,6 +388,13 @@ static HYPFS_STRING_INIT(compile_by, "compile_by"); static HYPFS_STRING_INIT(compile_date, "compile_date"); static HYPFS_STRING_INIT(compile_domain, "compile_domain"); static HYPFS_STRING_INIT(extra, "extra"); +static struct hypfs_entry_leaf config = { + .e.type =XEN_HYPFS_TYPE_STRING , + .e.encoding =XEN_HYPFS_ENC_GZIP , + .e.name ="config" , + .e.read = hypfs_read_leaf, + .content = &xen_config_data +};static int __init buildinfo_init (void ) {
@@ -414,6 +421,9 @@ static int __init buildinfo_init(void)hypfs_add_leaf (&version, &major, true);hypfs_add_leaf (&version, &minor, true); + config.e.size = xen_config_data_size; +hypfs_add_leaf (&buildinfo, &config, true); +return 0 ; }__initcall (buildinfo_init);
xen/include/xen/kernel.h +3 -0
@@ -100,5 +100,8 @@ extern enum system_state {bool_t is_active_kernel_text (unsigned long addr); +extern char xen_config_data; +extern unsigned int xen_config_data_size; +#endif /* _LINUX_KERNEL_H */
Signature · 7 lines
2.16.4 _______________________________________________ Xen-devel mailing list [email protected] https://lists.xenproject.org/mailman/listinfo/xen-devel
PATCH v3 9/9 xen: add runtime parameter access support to hypfs
In reply to Juergen GrossAdd support to read and modify values of hypervisor runtime parameters via the hypervisor file system.
As runtime parameters can be modified via a sysctl, too, this path has to take the hypfs rw_lock as writer.
For custom runtime parameters the resulting parameter value needs to be stored in a string buffer for being consumable by the file system.
---
V3:
- complete rework
- support custom parameters, too
- support parameter writing
--- docs/misc/hypfs-paths.pandoc | 9 ++++++ xen/arch/arm/xen.lds.S | 5 +++ xen/arch/x86/hvm/vmx/vmcs.c | 17 +++++++++- xen/arch/x86/pv/domain.c | 7 ++++- xen/arch/x86/xen.lds.S | 5 +++ xen/common/grant_table.c | 17 +++++++--- xen/common/hypfs.c | 38 +++++++++++++++++++++++ xen/common/kernel.c | 23 +++++++++++++- xen/drivers/char/console.c | 49 +++++++++++++++++++++++++---- xen/include/xen/hypfs.h | 4 +++ xen/include/xen/param.h | 74 +++++++++++++++++++++++++++++++++++++++----- 11 files changed, 226 insertions(+), 22 deletions(-)
docs/misc/hypfs-paths.pandoc +9 -0
@@ -142,3 +142,12 @@ The major version of Xen. #### /buildinfo/version/minor = INTEGER The minor version of Xen. + +#### /params/ + +A directory of runtime parameters. + +#### /params/* + +The single parameters. The description of the different parameters can be +found in `docs/misc/xen-command-line.pandoc`.
xen/arch/arm/xen.lds.S +5 -0
@@ -89,6 +89,11 @@ SECTIONS __start_schedulers_array = .; *(.data.schedulers) __end_schedulers_array = .; + + . = ALIGN(8); + __paramhypfs_start = .; + *(.data.paramhypfs) + __paramhypfs_end = .; *(.data.rel) *(.data.rel.*) CONSTRUCTORS
xen/arch/x86/hvm/vmx/vmcs.c +16 -1
@@ -70,6 +70,17 @@ integer_param("ple_window", ple_window);static bool __read_mostly opt_ept_pml = true;static s8 __read_mostly opt_ept_ad =-1 ;int8_t __read_mostly opt_ept_exec_sp =-1 ; +static char opt_ept_setting[16 ] ="pml=1" ; + + +static void update_ept_param (void ) +{ +snprintf (opt_ept_setting,sizeof (opt_ept_setting),"pml=%d" , opt_ept_pml); +if ( opt_ept_ad >=0 ) +param_append_str (opt_ept_setting,",ad=%d" , opt_ept_ad); +if ( opt_ept_exec_sp >=0 ) +param_append_str (opt_ept_setting,",exec-sp=%d" , opt_ept_exec_sp); +}static int __init parse_ept_param (const char *s) {
@@ -93,6 +104,8 @@ static int __init parse_ept_param(const char *s) s = ss +1 ; }while ( *ss ); +update_ept_param (); +return rc; }custom_param ("ept" , parse_ept_param);
@@ -115,6 +128,8 @@ static int parse_ept_param_runtime(const char *s) opt_ept_exec_sp = val; +update_ept_param (); +rcu_read_lock (&domlist_read_lock);for_each_domain ( d ) {
@@ -144,7 +159,7 @@ static int parse_ept_param_runtime(const char *s)return 0 ; } -custom_runtime_only_param ("ept" , parse_ept_param_runtime); +custom_runtime_only_param ("ept" , parse_ept_param_runtime, opt_ept_setting);/* Dynamic (run-time adjusted) execution control flags. */ u32 vmx_pin_based_exec_control__read_mostly ;
xen/arch/x86/pv/domain.c +6 -1
@@ -22,6 +22,7 @@ static __read_mostly enum {PCID_XPTI , PCID_NOXPTI } opt_pcid =PCID_XPTI ; +static char opt_pcid_val[7 ] ="xpti" ;static int parse_pcid (const char *s) {
@@ -31,10 +32,12 @@ static int parse_pcid(const char *s) {case 0 : opt_pcid =PCID_OFF ; +snprintf (opt_pcid_val,sizeof (opt_pcid_val),"off" );break ;case 1 : opt_pcid =PCID_ALL ; +snprintf (opt_pcid_val,sizeof (opt_pcid_val),"on" );break ;default :
@@ -42,10 +45,12 @@ static int parse_pcid(const char *s) {case 0 : opt_pcid =PCID_NOXPTI ; +snprintf (opt_pcid_val,sizeof (opt_pcid_val),"noxpti" );break ;case 1 : opt_pcid =PCID_XPTI ; +snprintf (opt_pcid_val,sizeof (opt_pcid_val),"xpti" );break ;default :
@@ -57,7 +62,7 @@ static int parse_pcid(const char *s)return rc; } -custom_runtime_param ("pcid" , parse_pcid); +custom_runtime_param ("pcid" , parse_pcid, opt_pcid_val);static void noreturncontinue_nonidle_domain (struct vcpu *v) {
xen/arch/x86/xen.lds.S +5 -0
@@ -280,6 +280,11 @@ SECTIONS __start_schedulers_array = .; *(.data.schedulers) __end_schedulers_array = .; + + . = ALIGN(8); + __paramhypfs_start = .; + *(.data.paramhypfs) + __paramhypfs_end = .; } :text DECL_SECTION(.data) {
xen/common/grant_table.c +12 -5
@@ -85,8 +85,10 @@ struct grant_table {struct grant_table_arch arch; }; +#define PAR_VAL_SZ 12 +static int parse_gnttab_limit (const char *param,const char *arg, -unsigned int *valp) +unsigned int *valp,char *par_val) {const char *e;unsigned long val;
@@ -99,28 +101,33 @@ static int parse_gnttab_limit(const char *param, const char *arg,return -ERANGE ; *valp = val; +snprintf (par_val,PAR_VAL_SZ ,"%lu" , val);return 0 ; }unsigned int __read_mostly opt_max_grant_frames =64 ; +static char gnttab_max_frames_val[PAR_VAL_SZ ] ="64" ;static int parse_gnttab_max_frames (const char *arg) {return parse_gnttab_limit ("gnttab_max_frames" , arg, - &opt_max_grant_frames); + &opt_max_grant_frames, gnttab_max_frames_val); } -custom_runtime_param ("gnttab_max_frames" , parse_gnttab_max_frames); +custom_runtime_param ("gnttab_max_frames" , parse_gnttab_max_frames, + gnttab_max_frames_val);static unsigned int __read_mostly opt_max_maptrack_frames =1024 ; +static char max_maptrack_frames_val[PAR_VAL_SZ ] ="1024" ;static int parse_gnttab_max_maptrack_frames (const char *arg) {return parse_gnttab_limit ("gnttab_max_maptrack_frames" , arg, - &opt_max_maptrack_frames); + &opt_max_maptrack_frames, + max_maptrack_frames_val); }custom_runtime_param ("gnttab_max_maptrack_frames" , - parse_gnttab_max_maptrack_frames); + parse_gnttab_max_maptrack_frames, max_maptrack_frames_val);#ifndef GNTTAB_MAX_VERSION #define GNTTAB_MAX_VERSION 2
xen/common/hypfs.c +38 -0
@@ -10,6 +10,7 @@#include <xen/guest_access.h> #include <xen/hypercall.h> #include <xen/hypfs.h> +#include <xen/param.h> #include <xen/rwlock.h> #include <public/hypfs.h>
@@ -296,6 +297,33 @@ int hypfs_write_bool(struct hypfs_entry_leaf *leaf,return 0 ; } +int hypfs_write_custom (struct hypfs_entry_leaf *leaf, +XEN_GUEST_HANDLE_PARAM (void ) uaddr,unsigned long ulen) +{ +struct param_hypfs *p; +char *buf; +int ret; + + buf =xzalloc_array (char, ulen); +if ( !buf ) +return -ENOMEM ; + + ret = -EFAULT ; +if (copy_from_guest (buf, uaddr, ulen) ) + goto out; + + ret = -EDOM ; +if ( buf[ulen -1 ] ) + goto out; + + p = container_of(leaf, struct param_hypfs, hypfs); + ret = p->param->par.func (buf); + + out: +xfree (buf); +return ret; +} +static int hypfs_write (struct hypfs_entry *entry,XEN_GUEST_HANDLE_PARAM (void ) uaddr,unsigned long ulen) {
@@ -363,3 +391,13 @@ long do_hypfs_op(unsigned int cmd,return ret; } + +void hypfs_write_lock (void ) +{ +write_lock (&hypfs_lock); +} + +void hypfs_write_unlock (void ) +{ +write_unlock (&hypfs_lock); +}
xen/common/kernel.c +22 -1
@@ -198,7 +198,13 @@ static void __init _cmdline_parse(const char *cmdline)int runtime_parse (const char *line) { -return parse_params (line, __param_start, __param_end); +int ret; + +hypfs_write_lock (); + ret =parse_params (line, __param_start, __param_end); +hypfs_write_unlock (); + +return ret; }/**
@@ -428,6 +434,21 @@ static int __init buildinfo_init(void) }__initcall (buildinfo_init); +static HYPFS_DIR_INIT(params, "params"); + +static int __init param_init (void ) +{ +struct param_hypfs *param; + +hypfs_add_dir (&hypfs_root, ¶ms, true); + +for ( param = __paramhypfs_start; param < __paramhypfs_end; param++ ) +hypfs_add_leaf (¶ms, ¶m->hypfs, true); + +return 0 ; +} +__initcall (param_init); +# define DO (fn) long do_##fn#endif
xen/drivers/char/console.c +43 -6
136 lines not on this page · raw message
xen/include/xen/hypfs.h +4 -0
@@ -85,5 +85,9 @@ int hypfs_write_leaf(struct hypfs_entry_leaf *leaf, XEN_GUEST_HANDLE_PARAM(void) uaddr, unsigned long ulen);int hypfs_write_bool (struct hypfs_entry_leaf *leaf,XEN_GUEST_HANDLE_PARAM (void ) uaddr,unsigned long ulen); +int hypfs_write_custom (struct hypfs_entry_leaf *leaf, +XEN_GUEST_HANDLE_PARAM (void ) uaddr,unsigned long ulen); +void hypfs_write_lock (void ); +void hypfs_write_unlock (void );#endif /* __XEN_HYPFS_H__ */
xen/include/xen/param.h +66 -8
126 lines not on this page · raw message
Signature · 7 lines
2.16.4 _______________________________________________ Xen-devel mailing list [email protected] https://lists.xenproject.org/mailman/listinfo/xen-devel
Quoted 4 lines · Juergen Gross
On the 2019 Xen developer summit there was agreement that the Xen hypervisor should gain support for a hierarchical name-value store similar to the Linux kernel's sysfs.
Is there a short summary of the most recent use cases for this feature and expected interactions with other Xen features (e.g. Panopticon Xen, security controls on information that is visible to guests, e.g. recent discussion on version number hiding). This would impact many subsystems.
Presumably Kconfig could enable/disable this optional feature and all dependencies, and the Xen toolstack would continue to function normally in its absence.
Rich
_______________________________________________
Xen-devel mailing list
[email protected]
https://lists.xenproject.org/mailman/listinfo/xen-devel
Quoted 7 lines · Rich Persaud
Quoted 4 lines · Juergen Gross
On the 2019 Xen developer summit there was agreement that the Xen hypervisor should gain support for a hierarchical name-value store similar to the Linux kernel's sysfs.
Is there a short summary of the most recent use cases for this feature and expected interactions with other Xen features (e.g. Panopticon Xen, security controls on information that is visible to guests, e.g. recent discussion on version number hiding). This would impact many subsystems.
In the first run access is permitted to dom0 only. Access to other guests needs to be discussed.
Current use cases are just the buildinfo leafs including the .config of the hypervisor, plus reading and writing runtime parameters.
I'd like to add per-cpupool parameters (like SMT per cpupool, scheduling granularity) and maybe per-domain ones (e.g. mitigation settings). Another area to cover would be debugging interfaces like lock profiling, performance counters, ...
Quoted 1 line
Presumably
Kconfigcould enable/disable this optional feature and all dependencies, and the Xen toolstack would continue to function normally in its absence.
I'd rather go the other way round: have a detailed look which current privileged interfaces (domctl, sysctl) can be replaced by the file system and switch over to it with (where necessary) fine grained access control. I think this is something to discuss at the next Xen summit in summer (I have already registered a session for that purpose).
Juergen
_______________________________________________
Xen-devel mailing list
[email protected]
https://lists.xenproject.org/mailman/listinfo/xen-devel