Skip to contentSkip to conversation

tools/tests: More cleanup for automation improvements

PATCH Superseded Andrew Cooper 2 participants 16 messages Jun 15, 2021 – Jun 16, 2021

v1 of 2 Superseded Latest is v2 →

lore
Share & tools
lore
#0 lore raw

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
#1 lore raw

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
index 38a085e398..d4b90303b2 100644
@@ -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
index ce6f937d0c..73ce95e6d7 100644
@@ -1,4 +1,5 @@
 xen-access
+xen-mceinj
 xen-memshare
 xen-ucode
 xen-vmtrace
tools/misc/Makefile +4 -0
index 2b683819d4..1a07191d83 100644
@@ -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
similarity index 97%
rename from tools/tests/mce-test/tools/xen-mceinj.c
rename to tools/misc/xen-mceinj.c
index 1187d01e5f..df55eefbac 100644
@@ -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
index 8746aabe6b..f76feb0b3a 100644
@@ -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
deleted file mode 100644
index 1395df38ac..0000000000
@@ -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
deleted file mode 100644
index 05f754dfd6..0000000000
@@ -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
deleted file mode 100644
index 0e92ac2977..0000000000
@@ -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) $(LDLIBS_libxenstore)
tools/tests/mce-test/tools/README deleted +0 -24
deleted file mode 100644
index bd0d442bae..0000000000
@@ -1,24 +0,0 @@
-Xen Machine Check Exception(MCE) error inject tool
-----------------------------------------------
-
-xen-mceinj is a software MCE injection tool, which is based on Xen 
-MCE injection mechanism. It allows to inject machine check errors on the
-software level into a running Xen/dom0/VM. This is intended for
-validation of the Xen machine check handler.
-
-With the help of the Makefile, it is possible to compile a binary file 
-named "xen-mceinj".
-
-Usage
------
-$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
Acked-by on v1 1/5
#2 lore raw

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.

While I'm not generally in favor of dropping testing code, given the constraints you mention

Jan

#3 lore raw

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.

While I'm not generally in favor of dropping testing code, given the constraints you mention

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

#4 lore raw

PATCH 2/5 tools/tests: Drop run runes

In reply to Andrew Cooper

These 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
index 70ff154da6..53b4f28b2a 100644
@@ -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
index 4bef482966..da5e2a4f9b 100644
@@ -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
index 5075bc2be2..f172cefd3d 100644
@@ -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
index 7b07c31bbd..7b3f58b7a2 100644
@@ -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
#5 lore raw

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

#6 lore raw

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/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.

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

#7 lore raw

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/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.

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

#8 lore raw

PATCH 3/5 tests/resource: Rework Makefile

In reply to 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.

---
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
index da5e2a4f9b..b22eb6fc21 100644
@@ -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
#9 lore raw

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.

---
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
index da5e2a4f9b..b22eb6fc21 100644

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

#10 lore raw

PATCH 4/5 tests/cpu-policy: Rework Makefile

In reply to Andrew Cooper

In 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
index 53b4f28b2a..ab3e0ffde3 100644
@@ -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
#11 lore raw

PATCH 5/5 tests/xenstore: Rework Makefile

In reply to 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_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
index d4b90303b2..8ebb51b6c5 100644
@@ -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
new file mode 100644
index 0000000000..4b44f5dd60
@@ -0,0 +1 @@
+test-xenstore
tools/tests/xenstore/Makefile +22 -9
index a367d88803..3c7b7073fd 100644
@@ -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
similarity index 100%
rename from tools/tests/xenstore/xs-test.c
rename to tools/tests/xenstore/test-xenstore.c
Signature · 1 line
2.11.0
#12 lore raw

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_FLAGS too.

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

#13 lore raw
In reply to Andrew Cooper
Quoted 4 lines · Andrew Cooper

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)?

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

#14 lore raw
Quoted 11 lines · Jan Beulich
Quoted 4 lines · Andrew Cooper

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)?

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

#15 lore raw
Quoted 21 lines · Andrew Cooper
Quoted 11 lines · Jan Beulich
Quoted 4 lines · Andrew Cooper

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)?

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_emulator in 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