tools/tests: More cleanup for automation improvements
cover tools/tests: More cleanup for automation improvements
This series takes us one step closer towards the "autotests" plan for simplifying the running of unit and low level system tests.
With this series in place, all the tests which need running in dom0 have working install/uninstall targets, so they can be packaged suitably.
Some questions, concerning future changes.
Ian: Presuambly OSSTest is going to complain if I stick a test- prefix on depriv-fd-checker? I've left that test alone for now, as it was the only preexisting test with working install runes.
Jan/Roger: x86_emulator and vpci use $(HOSTCC) not $(CC). While they are unit tests, we still potentially want to run them in dom0 rather than the build environment - particularly for x86_emulator which is heavily CPUID based and wants to run on a wide set of hardware. Any issues moving them off $(HOSTCC)?
Roger: vhpet isn't even wired into the build system, and seems non-trivial to run in the first place. How should "success" be judged?
Andrew Cooper (5):
tools/tests: Drop obsolete mce-test infrastructure tools/tests: Drop run runes tests/resource: Rework Makefile tests/cpu-policy: Rework Makefile tests/xenstore: Rework Makefile
.gitignore | 2 -
tools/misc/.gitignore | 1 +
tools/misc/Makefile | 4 +
tools/{tests/mce-test/tools => misc}/xen-mceinj.c | 32 +--
tools/tests/Makefile | 1 -
tools/tests/cpu-policy/Makefile | 33 +--
tools/tests/mce-test/Makefile | 12 -
tools/tests/mce-test/README | 75 ------
tools/tests/mce-test/cases/srao_llc/dom0/cases.sh | 73 ------
tools/tests/mce-test/cases/srao_llc/guest/cases.sh | 94 --------
tools/tests/mce-test/cases/srao_llc/xen/cases.sh | 69 ------
tools/tests/mce-test/cases/srao_mem/dom0/cases.sh | 73 ------
tools/tests/mce-test/cases/srao_mem/guest/cases.sh | 94 --------
tools/tests/mce-test/cases/srao_mem/xen/cases.sh | 69 ------
tools/tests/mce-test/cases/ucna_llc/dom0/cases.sh | 72 ------
tools/tests/mce-test/cases/ucna_llc/guest/cases.sh | 92 --------
tools/tests/mce-test/cases/ucna_llc/xen/cases.sh | 68 ------
tools/tests/mce-test/config/setup.conf | 24 --
tools/tests/mce-test/lib/xen-mceinj-tool.sh | 260 ---------------------
tools/tests/mce-test/tools/Makefile | 24 --
tools/tests/mce-test/tools/README | 24 --
tools/tests/resource/Makefile | 11 +-
tools/tests/vpci/Makefile | 4 -
tools/tests/x86_emulator/Makefile | 4 -
tools/tests/xenstore/.gitignore | 1 +
tools/tests/xenstore/Makefile | 31 ++-
.../tests/xenstore/{xs-test.c => test-xenstore.c} | 0
27 files changed, 71 insertions(+), 1176 deletions(-)
rename tools/{tests/mce-test/tools => misc}/xen-mceinj.c (97%)
delete mode 100644 tools/tests/mce-test/Makefile
delete mode 100644 tools/tests/mce-test/README
delete mode 100644 tools/tests/mce-test/cases/srao_llc/dom0/cases.sh
delete mode 100644 tools/tests/mce-test/cases/srao_llc/guest/cases.sh
delete mode 100644 tools/tests/mce-test/cases/srao_llc/xen/cases.sh
delete mode 100644 tools/tests/mce-test/cases/srao_mem/dom0/cases.sh
delete mode 100644 tools/tests/mce-test/cases/srao_mem/guest/cases.sh
delete mode 100644 tools/tests/mce-test/cases/srao_mem/xen/cases.sh
delete mode 100644 tools/tests/mce-test/cases/ucna_llc/dom0/cases.sh
delete mode 100644 tools/tests/mce-test/cases/ucna_llc/guest/cases.sh
delete mode 100644 tools/tests/mce-test/cases/ucna_llc/xen/cases.sh
delete mode 100644 tools/tests/mce-test/config/setup.conf
delete mode 100644 tools/tests/mce-test/lib/xen-mceinj-tool.sh
delete mode 100644 tools/tests/mce-test/tools/Makefile
delete mode 100644 tools/tests/mce-test/tools/README
create mode 100644 tools/tests/xenstore/.gitignore
rename tools/tests/xenstore/{xs-test.c => test-xenstore.c} (100%)Signature · 1 line
2.11.0
PATCH 1/5 tools/tests: Drop obsolete mce-test infrastructure
mce-test has a test suite, but it depends on xend, needs to run in-tree, and requires manual setup of at least one guest, and manual parameters to pass into cases. Drop the test infrasturcture.
Move the one useful remaining item, xen-mceinj, into misc/, fixing some minor style issues as it goes.
---
CC: Ian Jackson <[email protected]>
CC: Wei Liu <[email protected]>
CC: Jan Beulich <[email protected]>
CC: Roger Pau Monné <[email protected]>
CC: Juergen Gross <[email protected]>
---
.gitignore | 1 -
tools/misc/.gitignore | 1 +
tools/misc/Makefile | 4 +
tools/{tests/mce-test/tools => misc}/xen-mceinj.c | 32 +--
tools/tests/Makefile | 1 -
tools/tests/mce-test/Makefile | 12 -
tools/tests/mce-test/README | 75 ------
tools/tests/mce-test/cases/srao_llc/dom0/cases.sh | 73 ------
tools/tests/mce-test/cases/srao_llc/guest/cases.sh | 94 --------
tools/tests/mce-test/cases/srao_llc/xen/cases.sh | 69 ------
tools/tests/mce-test/cases/srao_mem/dom0/cases.sh | 73 ------
tools/tests/mce-test/cases/srao_mem/guest/cases.sh | 94 --------
tools/tests/mce-test/cases/srao_mem/xen/cases.sh | 69 ------
tools/tests/mce-test/cases/ucna_llc/dom0/cases.sh | 72 ------
tools/tests/mce-test/cases/ucna_llc/guest/cases.sh | 92 --------
tools/tests/mce-test/cases/ucna_llc/xen/cases.sh | 68 ------
tools/tests/mce-test/config/setup.conf | 24 --
tools/tests/mce-test/lib/xen-mceinj-tool.sh | 260 ---------------------
tools/tests/mce-test/tools/Makefile | 24 --
tools/tests/mce-test/tools/README | 24 --
20 files changed, 24 insertions(+), 1138 deletions(-)
rename tools/{tests/mce-test/tools => misc}/xen-mceinj.c (97%)
delete mode 100644 tools/tests/mce-test/Makefile
delete mode 100644 tools/tests/mce-test/README
delete mode 100644 tools/tests/mce-test/cases/srao_llc/dom0/cases.sh
delete mode 100644 tools/tests/mce-test/cases/srao_llc/guest/cases.sh
delete mode 100644 tools/tests/mce-test/cases/srao_llc/xen/cases.sh
delete mode 100644 tools/tests/mce-test/cases/srao_mem/dom0/cases.sh
delete mode 100644 tools/tests/mce-test/cases/srao_mem/guest/cases.sh
delete mode 100644 tools/tests/mce-test/cases/srao_mem/xen/cases.sh
delete mode 100644 tools/tests/mce-test/cases/ucna_llc/dom0/cases.sh
delete mode 100644 tools/tests/mce-test/cases/ucna_llc/guest/cases.sh
delete mode 100644 tools/tests/mce-test/cases/ucna_llc/xen/cases.sh
delete mode 100644 tools/tests/mce-test/config/setup.conf
delete mode 100644 tools/tests/mce-test/lib/xen-mceinj-tool.sh
delete mode 100644 tools/tests/mce-test/tools/Makefile
delete mode 100644 tools/tests/mce-test/tools/README.gitignore +0 -1
@@ -276,7 +276,6 @@ tools/tests/x86_emulator/test_x86_emulator tools/tests/x86_emulator/x86_emulate tools/tests/x86_emulator/xop*.[ch] tools/tests/xenstore/xs-test -tools/tests/mce-test/tools/xen-mceinj tools/tests/vpci/list.h tools/tests/vpci/vpci.[hc] tools/tests/vpci/test_vpci
tools/misc/.gitignore +1 -0
@@ -1,4 +1,5 @@ xen-access +xen-mceinj xen-memshare xen-ucode xen-vmtrace
tools/misc/Makefile +4 -0
@@ -22,6 +22,7 @@ INSTALL_SBIN-$(CONFIG_MIGRATE) += xen-hptool INSTALL_SBIN-$(CONFIG_X86) += xen-hvmcrash INSTALL_SBIN-$(CONFIG_X86) += xen-hvmctx INSTALL_SBIN-$(CONFIG_X86) += xen-lowmemd +INSTALL_SBIN-$(CONFIG_X86) += xen-mceinj INSTALL_SBIN-$(CONFIG_X86) += xen-memshare INSTALL_SBIN-$(CONFIG_X86) += xen-mfndump INSTALL_SBIN-$(CONFIG_X86) += xen-ucode
@@ -97,6 +98,9 @@ xen-memshare: xen-memshare.o xen-vmtrace: xen-vmtrace.o $(CC) $(LDFLAGS) -o $@ $< $(LDLIBS_libxenctrl) $(LDLIBS_libxenforeignmemory) $(APPEND_LDFLAGS) +xen-mceinj: xen-mceinj.o + $(CC) $(LDFLAGS) -o $@ $< $(LDLIBS_libxenctrl) $(LDLIBS_libxenguest) $(LDLIBS_libxenstore) $(APPEND_LDFLAGS) + xenperf: xenperf.o $(CC) $(LDFLAGS) -o $@ $< $(LDLIBS_libxenctrl) $(APPEND_LDFLAGS)
tools/misc/xen-mceinj.c renamed from tools/tests/mce-test/tools/xen-mceinj.c +19 -13
@@ -137,7 +137,7 @@ static void err(xc_interface *xc_handle, const char *fmt, ...)va_list args;va_start (args, fmt); -if (vasprintf (&buf, fmt, args) <0 ) +if (vasprintf (&buf, fmt, args) <0 )abort ();perror (buf);va_end (args);
@@ -173,7 +173,7 @@ static unsigned int mca_cpuinfo(xc_interface *xc_handle) mc.cmd = XEN_MC_physcpuinfo; mc.interface_version =XEN_MCA_INTERFACE_VERSION ; -if (!xc_mca_op (xc_handle, &mc)) +if ( !xc_mca_op (xc_handle, &mc) )return mc.u.mc_physcpuinfo.ncpus;else return 0 ;
@@ -187,9 +187,9 @@ static int inject_cmci(xc_interface *xc_handle, unsigned int cpu_nr)memset (&mc,0 ,sizeof (struct xen_mc )); nr_cpus =mca_cpuinfo (xc_handle); -if (!nr_cpus) +if ( !nr_cpus )err (xc_handle,"Failed to get mca_cpuinfo" ); -if (cpu_nr >= nr_cpus) +if ( cpu_nr >= nr_cpus )err (xc_handle,"-c %u is larger than %u" , cpu_nr, nr_cpus -1 ); mc.cmd = XEN_MC_inject_v2;
@@ -284,7 +284,7 @@ static int add_msr_intpose(xc_interface *xc_handle,flush_msr_inj (xc_handle);init_msr_inj (); } - count= msr_inj.mcinj_count; + count = msr_inj.mcinj_count;if ( !count ) {
@@ -422,7 +422,7 @@ static long xs_get_dom_mem(int domid)if (!memstr || !plen)return -1 ; - mem =atoll (memstr)*1024 ; + mem =atoll (memstr) *1024 ;free (memstr);return mem;
@@ -474,17 +474,20 @@ int main(int argc, char *argv[]) cpu_nr =0 ;init_msr_inj (); - xc_handle =xc_interface_open (0 ,0 ,0 ); -if ( !xc_handle ) { + xc_handle =xc_interface_open (NULL ,NULL ,0 ); +if ( !xc_handle ) + {Lprintf ("Failed to get xc interface" );exit (EXIT_FAILURE ); } -while (1 ) { +while (1 ) + { c =getopt_long (argc, argv,"c:Dd:t:hp:l" , opts, &opt_index);if ( c ==-1 )break ; -switch ( c ) { +switch ( c ) + {case 'D' : dump=1 ;break ;
@@ -516,7 +519,8 @@ int main(int argc, char *argv[]) } } -if ( domid !=DOMID_XEN ) { +if ( domid !=DOMID_XEN ) + { max_gpa =xs_get_dom_mem (domid);Lprintf ("get domain %d max gpa is: 0x%lx" , domid, max_gpa);if ( gaddr >= max_gpa )
@@ -524,7 +528,8 @@ int main(int argc, char *argv[]) }Lprintf ("get gaddr of error inject is: 0x%lx" , gaddr); -if ( dump ) { +if ( dump ) + {if ( domid ==DOMID_XEN )Lprintf ("Xen: gaddr=0x%lx" , gaddr);else
@@ -532,7 +537,8 @@ int main(int argc, char *argv[]) goto out; } -if ( type <0 || type >=MCE_TABLE_SIZE ) { +if ( type <0 || type >=MCE_TABLE_SIZE ) + {err (xc_handle,"Unsupported error type" ); goto out; }
tools/tests/Makefile +0 -1
@@ -4,7 +4,6 @@ include $(XEN_ROOT)/tools/Rules.mk SUBDIRS-y := SUBDIRS-y += resource SUBDIRS-$(CONFIG_X86) += cpu-policy -SUBDIRS-$(CONFIG_X86) += mce-test ifneq ($(clang),y) SUBDIRS-$(CONFIG_X86) += x86_emulator endif
tools/tests/mce-test/Makefile deleted +0 -12
@@ -1,12 +0,0 @@ -.PHONY: all clean distclean - -all: - $(MAKE) -C tools - -clean: - $(MAKE) -C tools clean - -distclean: - $(MAKE) -C tools distclean - -install uninstall:
tools/tests/mce-test/README deleted +0 -75
75 lines not on this page · raw message
tools/tests/mce-test/cases/srao_llc/dom0/cases.sh deleted +0 -73
73 lines not on this page · raw message
tools/tests/mce-test/cases/srao_llc/guest/cases.sh deleted +0 -94
94 lines not on this page · raw message
tools/tests/mce-test/cases/srao_llc/xen/cases.sh deleted +0 -69
69 lines not on this page · raw message
tools/tests/mce-test/cases/srao_mem/dom0/cases.sh deleted +0 -73
73 lines not on this page · raw message
tools/tests/mce-test/cases/srao_mem/guest/cases.sh deleted +0 -94
94 lines not on this page · raw message
tools/tests/mce-test/cases/srao_mem/xen/cases.sh deleted +0 -69
69 lines not on this page · raw message
tools/tests/mce-test/cases/ucna_llc/dom0/cases.sh deleted +0 -72
72 lines not on this page · raw message
tools/tests/mce-test/cases/ucna_llc/guest/cases.sh deleted +0 -92
92 lines not on this page · raw message
tools/tests/mce-test/cases/ucna_llc/xen/cases.sh deleted +0 -68
68 lines not on this page · raw message
tools/tests/mce-test/config/setup.conf deleted +0 -24
@@ -1,24 +0,0 @@ -#!/bin/bash -# -# Software injection based test cases: test cases are triggered via -# mce-inject tool. -# Copyright (c) 2010, Intel Corporation -# -# This program is free software; you can redistribute it and/or -# modify it under the terms of the GNU General Public License version -# 2 as published by the Free Software Foundation. -# -# This program is distributed in the hope that it will be useful, but -# WITHOUT ANY WARRANTY; without even the implied warranty of -# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU -# General Public License for more details. -# -# You should have received a copy of the GNU General Public License -# along with this program; If not, see <http://www.gnu.org/licenses/>. -# -# Author: Xudong Hao <[email protected]> -# - -export MCE_SRAO_MEM=0 -export MCE_SRAO_LLC=1 -export CMCI_UCNA_LLC=2
tools/tests/mce-test/lib/xen-mceinj-tool.sh deleted +0 -260
260 lines not on this page · raw message
tools/tests/mce-test/tools/Makefile deleted +0 -24
@@ -1,24 +0,0 @@ -XEN_ROOT=$(CURDIR)/../../../.. -include $(XEN_ROOT)/tools/Rules.mk - -CFLAGS += -Werror -CFLAGS += $(CFLAGS_libxenctrl) -CFLAGS += $(CFLAGS_libxenguest) -CFLAGS += $(CFLAGS_libxenstore) -CFLAGS += $(CFLAGS_xeninclude) - -.PHONY: all -all: xen-mceinj - -install: xen-mceinj - $(INSTALL_PROG) xen-mceinj $(DESTDIR)$(sbindir) - -.PHONY: clean -clean: - $(RM) *.o xen-mceinj - -.PHONY: distclean -distclean: clean - -xen-mceinj: xen-mceinj.o Makefile - $(CC) -o $@ $< $(LDFLAGS) $(LDLIBS_libxenctrl) $(LDLIBS_libxenguest) $(LDLIBS_libxenstore)
tools/tests/mce-test/tools/README deleted +0 -24
@@ -1,24 +0,0 @@ -Xen Machine Check Exception(MCE) error inject tool ----------------------------------------------- - -xen-mceinj is a software MCE injection tool, which is based on Xen -MCE injection mechanism. It allows to inject machine check errors on the -software level into a running Xen/dom0/VM. This is intended for -validation of the Xen machine check handler. - -With the help of the Makefile, it is possible to compile a binary file -named "xen-mceinj". - -Usage ------ -$make (make install) --Note: make sure compile xen/tools before do this step -$./xen-mceinj [OPTION]... - -OPTION arguments can be: - -D, --dump dump addr info without error injection - -c, --cpu=CPU_ID target CPU, the default is CPU0 - -d, --domain=DomID target domain, the default is Xen itself - -p, --page physical page address, the default is 0x180020 - -t, --type=error error type - -For detail help, please refer to "./xen-mceinj -h"
Signature · 1 line
2.11.0
Re: [PATCH 1/5] tools/tests: Drop obsolete mce-test infrastructure
Quoted 8 lines · Andrew Cooper
mce-test has a test suite, but it depends on xend, needs to run in-tree, and requires manual setup of at least one guest, and manual parameters to pass into cases. Drop the test infrasturcture.
Move the one useful remaining item, xen-mceinj, into misc/, fixing some minor style issues as it goes.
Signed-off-by Andrew Cooper <[email protected]>
While I'm not generally in favor of dropping testing code, given the constraints you mention
Jan
Re: [PATCH 1/5] tools/tests: Drop obsolete mce-test infrastructure
Quoted 5 lines · Jan Beulich
Quoted 4 lines · Andrew Cooper
mce-test has a test suite, but it depends on xend, needs to run in-tree, and requires manual setup of at least one guest, and manual parameters to pass into cases. Drop the test infrasturcture.
Quoted 7 lines
Quoted 4 lines
Move the one useful remaining item, xen-mceinj, into misc/, fixing some minor style issues as it goes.
Signed-off-by Andrew Cooper <[email protected]>While I'm not generally in favor of dropping testing code, given the constraints you mention
Acked-by Jan Beulich <[email protected]>
Thanks.
Testing at this level is really in the OSSTest/XenRT realm, as it involves coordination between guests and checking logs after the fact. However, the need to pass raw addresses around to make any of it work in the first place make this manual in practice.
~Andrew
PATCH 2/5 tools/tests: Drop run runes
In reply to Andrew CooperThese aren't implemented consistently. The one in resource/ is useless as the binary needs running in dom0, and the layout in cpu-policy/ demonstrates the weakness of this approach with multiple binaries.
---
CC: Ian Jackson <[email protected]>
CC: Wei Liu <[email protected]>
CC: Jan Beulich <[email protected]>
CC: Roger Pau Monné <[email protected]>
CC: Juergen Gross <[email protected]>
--- tools/tests/cpu-policy/Makefile | 4 ---- tools/tests/resource/Makefile | 4 ---- tools/tests/vpci/Makefile | 4 ---- tools/tests/x86_emulator/Makefile | 4 ---- 4 files changed, 16 deletions(-)
tools/tests/cpu-policy/Makefile +0 -4
@@ -17,10 +17,6 @@ endif .PHONY: all all: $(TARGET-y) -.PHONY: run -run: $(TARGET-y) - ./$(TARGET-y) - .PHONY: clean clean: $(RM) -f -- *.o .*.d .*.d2 test-cpu-policy
tools/tests/resource/Makefile +0 -4
@@ -6,10 +6,6 @@ TARGET := test-resource .PHONY: all all: $(TARGET) -.PHONY: run -run: $(TARGET) - ./$(TARGET) - .PHONY: clean clean: $(RM) -f -- *.o $(TARGET) $(DEPS_RM)
tools/tests/vpci/Makefile +0 -4
@@ -6,10 +6,6 @@ TARGET := test_vpci .PHONY: all all: $(TARGET) -.PHONY: run -run: $(TARGET) - ./$(TARGET) - $(TARGET): vpci.c vpci.h list.h main.c emul.h $(HOSTCC) -g -o $@ vpci.c main.c
tools/tests/x86_emulator/Makefile +0 -4
@@ -7,10 +7,6 @@ TARGET := test_x86_emulator .PHONY: all all: -.PHONY: run -run: $(TARGET) - ./$(TARGET) - # Add libx86 to the build vpath %.c $(XEN_ROOT)/xen/lib/x86
Signature · 1 line
2.11.0
Re: [PATCH 2/5] tools/tests: Drop run runes
tools/tests/x86_emulator/Makefile TARGET := test_x86_emulator · line 12
Quoted 13 lines · Andrew Cooper
tools/tests/x86_emulator/Makefile +0
-4@@ -7,10 +7,6 @@ TARGET := test_x86_emulator .PHONY: all all: -.PHONY: run -run: $(TARGET) - ./$(TARGET) - # Add libx86 to the build vpath %.c $(XEN_ROOT)/xen/lib/x86
This is not only incomplete, but actively (specifically here for my own frequent using of it, but other tests I do run occasionally as well, and then also that same way) harming things as long as you don't introduce an alternative way. Note the top-level Makefile making use of these rules, and note also the run32 companion here.
Jan
Re: [PATCH 2/5] tools/tests: Drop run runes
Quoted 7 lines · Jan Beulich
Quoted 6 lines · Andrew Cooper
tools/tests/x86_emulator/Makefile +0
-0@@ -7,10 +7,6 @@ TARGET := test_x86_emulator .PHONY: all all:
Quoted 12 lines
tools/tests/x86_emulator/Makefile TARGET := test_x86_emulator · line 12
Quoted 7 lines
tools/tests/x86_emulator/Makefile +0
-4-.PHONY: run -run: $(TARGET) - ./$(TARGET) - # Add libx86 to the build vpath %.c $(XEN_ROOT)/xen/lib/x86This is not only incomplete, but actively (specifically here for my own frequent using of it, but other tests I do run occasionally as well, and then also that same way) harming things as long as you don't introduce an alternative way. Note the top-level
Makefilemaking use of these rules, and note also the run32 companion here.
Honestly, this makefile is borderline impossible to follow. I failed to make the install runes work, which is part of why I deferred the unit-like tests for now.
But I'm taking this as a strong preference to keep the run target?
TBH, this patch is a little on the side of the rest of the series. I stand by my commit message though - these are inconsistent, and buggy in at least one case.
~Andrew
Re: [PATCH 2/5] tools/tests: Drop run runes
Quoted 21 lines · Andrew Cooper
Quoted 7 lines · Jan Beulich
Quoted 6 lines · Andrew Cooper
tools/tests/x86_emulator/Makefile +0
-0@@ -7,10 +7,6 @@ TARGET := test_x86_emulator .PHONY: all all:Quoted 12 lines
tools/tests/x86_emulator/Makefile TARGET := test_x86_emulator · line 12
Quoted 7 lines
tools/tests/x86_emulator/Makefile +0
-4-.PHONY: run -run: $(TARGET) - ./$(TARGET) - # Add libx86 to the build vpath %.c $(XEN_ROOT)/xen/lib/x86This is not only incomplete, but actively (specifically here for my own frequent using of it, but other tests I do run occasionally as well, and then also that same way) harming things as long as you don't introduce an alternative way. Note the top-level
Makefilemaking use of these rules, and note also the run32 companion here.
Quoted 3 lines
Honestly, this makefile is borderline impossible to follow. I failed to make the install runes work, which is part of why I deferred the unit-like tests for now.
Well, yes, it's not pretty, but I lack ideas for a clear improvement.
Quoted 1 line
But I'm taking this as a strong preference to keep the run target?
Yes, and not just here. Any test that can be run directly in the build tree would imo better have such a target, so re-building and running it can be invoked with a single make invocation. Having to re-build the entire tools/ in order to then (re-)run one of these tests is, in particular, way too much latency.
Jan
PATCH 3/5 tests/resource: Rework Makefile
In reply to Andrew CooperIn particular, fill in the install/uninstall rules so this test can be packaged to be automated sensibly.
Make all object files depend on the Makefile, and us $(TARGET) when appropriate.
---
CC: Ian Jackson <[email protected]>
CC: Wei Liu <[email protected]>
CC: Jan Beulich <[email protected]>
CC: Roger Pau Monné <[email protected]>
CC: Juergen Gross <[email protected]>
--- tools/tests/resource/Makefile | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-)
tools/tests/resource/Makefile +6 -1
@@ -16,9 +16,12 @@ distclean: clean .PHONY: install install: all + $(INSTALL_DIR) $(DESTDIR)$(LIBEXEC_BIN) + $(INSTALL_PROG) $(TARGET) $(DESTDIR)$(LIBEXEC_BIN) .PHONY: uninstall uninstall: + $(RM) -f -- $(DESTDIR)$(LIBEXEC_BIN)/$(TARGET) CFLAGS += -Werror CFLAGS += $(CFLAGS_xeninclude)
@@ -30,7 +33,9 @@ LDFLAGS += $(LDLIBS_libxenctrl) LDFLAGS += $(LDLIBS_libxenforeignmemory) LDFLAGS += $(APPEND_LDFLAGS) -test-resource: test-resource.o +*.o: Makefile + +$(TARGET): test-resource.o $(CC) -o $@ $< $(LDFLAGS) -include $(DEPS_INCLUDE)
Signature · 1 line
2.11.0
Re: [PATCH 3/5] tests/resource: Rework Makefile
Quoted 22 lines · Andrew Cooper
In particular, fill in the install/uninstall rules so this test can be packaged to be automated sensibly.
Make all object files depend on the
Makefile, and us $(TARGET) when appropriate.Signed-off-by Andrew Cooper <[email protected]>---
CC: Ian Jackson <[email protected]>
CC: Wei Liu <[email protected]>
CC: Jan Beulich <[email protected]>
CC: Roger Pau Monné <[email protected]>
CC: Juergen Gross <[email protected]>--- tools/tests/resource/Makefile | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-)tools/tests/resource/Makefile +0
-0
tools/tests/resource/Makefile distclean: clean · line 24
Quoted 9 lines
tools/tests/resource/Makefile +3
-0@@ -16,9 +16,12 @@ distclean: clean .PHONY: install install: all + $(INSTALL_DIR) $(DESTDIR)$(LIBEXEC_BIN) + $(INSTALL_PROG) $(TARGET) $(DESTDIR)$(LIBEXEC_BIN) .PHONY: uninstall uninstall: + $(RM) -f -- $(DESTDIR)$(LIBEXEC_BIN)/$(TARGET)
I've finally figured out where $(RM) comes from. Its a GNU Make default, and has -f included.
I've altered this and later patches in the series to take this into account.
~Andrew
PATCH 4/5 tests/cpu-policy: Rework Makefile
In reply to Andrew CooperIn particular, fill in the install/uninstall rules so this test can be packaged to be automated sensibly.
Rework TARGET-y to be TARGETS, drop the unconditional -O3 and use the default instead, and drop CFLAGS from the link line but honour APPEND_LDFLAGS.
---
CC: Ian Jackson <[email protected]>
CC: Wei Liu <[email protected]>
CC: Jan Beulich <[email protected]>
CC: Roger Pau Monné <[email protected]>
CC: Juergen Gross <[email protected]>
--- tools/tests/cpu-policy/Makefile | 29 ++++++++++++++++++----------- 1 file changed, 18 insertions(+), 11 deletions(-)
tools/tests/cpu-policy/Makefile +18 -11
@@ -1,25 +1,23 @@ XEN_ROOT = $(CURDIR)/../../.. include $(XEN_ROOT)/tools/Rules.mk -TARGET-y := test-cpu-policy +TARGETS := # For brevity, these tests make extensive use of designated initialisers in # anonymous unions, but GCCs older than 4.6 can't cope. Ignore the test in # this case. -ifneq ($(clang),y) -TARGET-$(call cc-ver,$(CC),lt,0x040600) := -endif - -ifeq ($(TARGET-y),) +ifneq ($(gcc)$(call cc-ver,$(CC),lt,0x040600),yy) +TARGETS += test-cpu-policy +else $(warning Test harness not built, use newer compiler than "$(CC)" (version $(shell $(CC) -dumpversion))) endif .PHONY: all -all: $(TARGET-y) +all: $(TARGETS) .PHONY: clean clean: - $(RM) -f -- *.o .*.d .*.d2 test-cpu-policy + $(RM) -f -- *.o $(TARGETS) $(DEPS_RM) .PHONY: distclean distclean: clean
@@ -27,15 +25,24 @@ distclean: clean .PHONY: install install: all + $(INSTALL_DIR) $(DESTDIR)$(LIBEXEC_BIN) + $(if $(TARGETS),$(INSTALL_PROG) $(TARGETS) $(DESTDIR)$(LIBEXEC_BIN)) .PHONY: uninstall +uninstall: + $(RM) -f -- $(addprefix $(DESTDIR)$(LIBEXEC_BIN)/,$(TARGETS)) -CFLAGS += -Werror $(CFLAGS_xeninclude) -D__XEN_TOOLS__ -O3 +CFLAGS += -Werror -D__XEN_TOOLS__ +CFLAGS += $(CFLAGS_xeninclude) CFLAGS += $(APPEND_CFLAGS) -vpath %.c ../../../xen/lib/x86 +LDFLAGS += $(APPEND_LDFLAGS) + +vpath %.c $(XEN_ROOT)/xen/lib/x86 + +*.o: Makefile test-cpu-policy: test-cpu-policy.o msr.o cpuid.o policy.o - $(CC) $(CFLAGS) $^ -o $@ + $(CC) $^ -o $@ $(LDFLAGS) -include $(DEPS_INCLUDE)
Signature · 1 line
2.11.0
PATCH 5/5 tests/xenstore: Rework Makefile
In reply to Andrew CooperIn particular, fill in the install/uninstall rules so this test can be packaged to be automated sensibly.
Rename xs-test to test-xenstore to be consistent with other tests. Honour APPEND_FLAGS too.
---
CC: Ian Jackson <[email protected]>
CC: Wei Liu <[email protected]>
CC: Jan Beulich <[email protected]>
CC: Roger Pau Monné <[email protected]>
CC: Juergen Gross <[email protected]>
---
.gitignore | 1 -
tools/tests/xenstore/.gitignore | 1 +
tools/tests/xenstore/Makefile | 31 +++++++++++++++-------
.../tests/xenstore/{xs-test.c => test-xenstore.c} | 0
4 files changed, 23 insertions(+), 10 deletions(-)
create mode 100644 tools/tests/xenstore/.gitignore
rename tools/tests/xenstore/{xs-test.c => test-xenstore.c} (100%).gitignore +0 -1
@@ -275,7 +275,6 @@ tools/tests/x86_emulator/*sse*.[ch] tools/tests/x86_emulator/test_x86_emulator tools/tests/x86_emulator/x86_emulate tools/tests/x86_emulator/xop*.[ch] -tools/tests/xenstore/xs-test tools/tests/vpci/list.h tools/tests/vpci/vpci.[hc] tools/tests/vpci/test_vpci
tools/tests/xenstore/.gitignore new file +1 -0
@@ -0,0 +1 @@ +test-xenstore
tools/tests/xenstore/Makefile +22 -9
@@ -1,11 +1,7 @@ XEN_ROOT=$(CURDIR)/../../.. include $(XEN_ROOT)/tools/Rules.mk -CFLAGS += -Werror - -CFLAGS += $(CFLAGS_libxenstore) - -TARGETS-y := xs-test +TARGETS-y := test-xenstore TARGETS := $(TARGETS-y) .PHONY: all
@@ -16,14 +12,31 @@ build: $(TARGETS) .PHONY: clean clean: - $(RM) *.o $(TARGETS) *~ $(DEPS_RM) + $(RM) -f -- *.o $(TARGETS) $(DEPS_RM) .PHONY: distclean distclean: clean + $(RM) -f -- *~ + +.PHONY: install +install: all + $(INSTALL_DIR) $(DESTDIR)$(LIBEXEC_BIN) + $(if $(TARGETS),$(INSTALL_PROG) $(TARGETS) $(DESTDIR)$(LIBEXEC_BIN)) + +.PHONY: uninstall +uninstall: + $(RM) -f -- $(addprefix $(DESTDIR)$(LIBEXEC_BIN)/,$(TARGETS)) + +CFLAGS += -Werror +CFLAGS += $(CFLAGS_libxenstore) +CFLAGS += $(APPEND_CFLAGS) + +LDFLAGS += $(LDLIBS_libxenstore) +LDFLAGS += $(APPEND_LDFLAGS) -xs-test: xs-test.o Makefile - $(CC) -o $@ $< $(LDFLAGS) $(LDLIBS_libxenstore) +%.o: Makefile -install uninstall: +test-xenstore: test-xenstore.o + $(CC) -o $@ $< $(LDFLAGS) -include $(DEPS_INCLUDE)
tools/tests/xenstore/test-xenstore.c renamed from tools/tests/xenstore/xs-test.c +0 -0
Signature · 1 line
2.11.0
Re: [PATCH 5/5] tests/xenstore: Rework Makefile
Quoted 7 lines · Andrew Cooper
In particular, fill in the install/uninstall rules so this test can be packaged to be automated sensibly.
Rename xs-test to test-xenstore to be consistent with other tests. Honour
APPEND_FLAGStoo.Signed-off-by Andrew Cooper <[email protected]>
It turns out that the absence of install rule previously meant that this wasn't covered by CI.
It triggers test-xenstore.c:486:5: error: ignoring return value of 'asprintf', declared with attribute warn_unused_result [-Werror=unused-result] on newer GCCs.
~Andrew
Quoted 4 lines · Andrew Cooper
Jan/Roger:
x86_emulatorand vpci use $(HOSTCC) not $(CC). While they are unit tests, we still potentially want to run them in dom0 rather than the build environment - particularly forx86_emulatorwhich is heavily CPUID based and wants to run on a wide set of hardware. Any issues moving them off $(HOSTCC)?
Well, yes, I'm afraid: If anything, we may need to build two binaries, or build the one binary two different ways: The "run" (and "run32" for the emulator harness) target wants a binary built with HOSTCC. The install target (which prior to your series does nothing) indeed wants building with CC. So maybe we want something like
install: HOSTCC:=$(CC)
plus suitable detection of whether the opposite set of objects are presently in the build area, requiring a full rebuild? (Of course this will work only as long as HOSTCC isn't used for any build time helper binaries. See "x86emul: test AMX insns" for when this starts not to be the case anymore for the emulator harness. So we'd need yet another variable to express this detail.)
Jan
Quoted 11 lines · Jan Beulich
Quoted 4 lines · Andrew Cooper
Jan/Roger:
x86_emulatorand vpci use $(HOSTCC) not $(CC). While they are unit tests, we still potentially want to run them in dom0 rather than the build environment - particularly forx86_emulatorwhich is heavily CPUID based and wants to run on a wide set of hardware. Any issues moving them off $(HOSTCC)?Well, yes, I'm afraid: If anything, we may need to build two binaries, or build the one binary two different ways: The "run" (and "run32" for the emulator harness) target wants a binary built with HOSTCC. The install target (which prior to your series does nothing) indeed wants building with CC. So maybe we want something like
Quoted 8 lines
install: HOSTCC:=$(CC)
plus suitable detection of whether the opposite set of objects are presently in the build area, requiring a full rebuild? (Of course this will work only as long as HOSTCC isn't used for any build time helper binaries. See "x86emul: test AMX insns" for when this starts not to be the case anymore for the emulator harness. So we'd need yet another variable to express this detail.)
Having slept on the problem overnight, I'm going to argue that HOSTCC is conceptually wrong to use here in the first place.
In an arm64 environment, cross-compiling x86_64, this will explode everywhere, and the fault is with using HOSTCC rather than CC.
HOSTCC is specifically for compiling utilities executed as part of the build. Tests, and particularly arch-specific ones like x86_emulate, are not in this category. Whether you happen to be able to run test_x86_emulator in the build environment is a property of whether you're cross-compiling.
For a non-cross-compiled builds, HOSTCC and CC are largely interchangeable, and won't impact the ability to run the binary in the build environment.
~Andrew
Quoted 21 lines · Andrew Cooper
Quoted 11 lines · Jan Beulich
Quoted 4 lines · Andrew Cooper
Jan/Roger:
x86_emulatorand vpci use $(HOSTCC) not $(CC). While they are unit tests, we still potentially want to run them in dom0 rather than the build environment - particularly forx86_emulatorwhich is heavily CPUID based and wants to run on a wide set of hardware. Any issues moving them off $(HOSTCC)?Well, yes, I'm afraid: If anything, we may need to build two binaries, or build the one binary two different ways: The "run" (and "run32" for the emulator harness) target wants a binary built with HOSTCC. The install target (which prior to your series does nothing) indeed wants building with CC. So maybe we want something like
Quoted 8 lines
install: HOSTCC:=$(CC)
plus suitable detection of whether the opposite set of objects are presently in the build area, requiring a full rebuild? (Of course this will work only as long as HOSTCC isn't used for any build time helper binaries. See "x86emul: test AMX insns" for when this starts not to be the case anymore for the emulator harness. So we'd need yet another variable to express this detail.)
Quoted 5 lines
Having slept on the problem overnight, I'm going to argue that HOSTCC is conceptually wrong to use here in the first place.
In an arm64 environment, cross-compiling
x86_64, this will explode everywhere, and the fault is with using HOSTCC rather than CC.
In principle, if there wasn't the massive amount of inline assembly, and if the emulator wasn't just re-executing the instructions it is asked to emulate, building all of this on Arm ought to be possible. But with the code we have this simply makes no sense.
Quoted 3 lines
HOSTCC is specifically for compiling utilities executed as part of the build. Tests, and particularly arch-specific ones like
x86_emulate, are not in this category.
Hmm, right now they definitely are. Running them directly (which is their only purpose right now, with the install targets doing nothing) from the build tree puts them into this category. But they aren't anymore as soon as you want to install them. Hence the need to have two modes here.
Quoted 3 lines
Whether you happen to be able to run
test_x86_emulatorin the build environment is a property of whether you're cross-compiling.
In a way, yes. I'd consider the run32 target to also be cross-like, yet that binary can then still be run from the build tree (if the distro supports 32-bit binaries).
Quoted 3 lines
For a non-cross-compiled builds, HOSTCC and CC are largely interchangeable, and won't impact the ability to run the binary in the build environment.
Not exactly, I think. If I override CC but not HOSTCC (which I do normally), I still don't need to worry about that other CC finding all the bits and pieces it needs for building a host binary. For the C compiler this may not mean much, but the C++ compiler (if we used it and hence would need to treat it similarly) wants e.g. its own copies of the headers, which may not be readily available. I will admit that my environment may be pretty non-standard, as I run non- default compilers (and binutils) also directly from their build trees. But things have been working fine this way for all sorts of projects, so I expect the test harness here to not break in this regard.
Jan