vmmon: fix indirect call with retpoline build

Build against kernel with retpoline support issues warning objtool: Task_Switch()+0x425: indirect call found in RETPOLINE build This is because an indirect call in TaskSwitchToMonitor() is encoded using inline assembler so that it bypasses retpoline generation. For this purpose, macro CALL_NOSPEC exists since v4.15-rc8 (and has been backported into some distribution kernels with the rest of retpoline support). Use the macro if available and fallback to the original code if not.
vmmon: fix always_inline attribute usage
2018-08-20 07:57:39 +02:00 · 2018-08-20 07:57:39 +02:00 · 2018-08-20 07:57:39 +02:00 · 2018-08-20 07:57:39 +02:00 · 2018-08-20 07:57:39 +02:00 · 2018-08-20 07:57:39 +02:00
5 changed files with 16 additions and 3 deletions
@@ -31,7 +31,7 @@ $(DRIVER)-y := $(subst $(SRCROOT)/, , $(patsubst %.c, %.o, \
 		$(SRCROOT)/bootstrap/*.c)))

 clean:
-	rm -rf $(wildcard $(DRIVER).mod.c $(DRIVER).ko .tmp_versions \
+	rm -rf $(wildcard $(DRIVER).mod.c $(DRIVER).ko .tmp_versions .cache.mk \
 		Module.symvers Modules.symvers Module.markers modules.order \
 		$(foreach dir,linux/ common/ vmcore/ bootstrap/ \
 		./,$(addprefix $(dir),.*.cmd .*.o.flags *.o)))
@@ -2203,12 +2203,23 @@ TaskSwitchToMonitor(VMCrossPage *crosspage)
   {
      uint64 raxGetsWiped, rcxGetsWiped;

+#ifdef CALL_NOSPEC
+      __asm__ __volatile__(CALL_NOSPEC
+                           : "=a" (raxGetsWiped),
+                             "=c" (rcxGetsWiped)
+                           : "0" (codePtr),
+                             "1" (crosspage),
+                             THUNK_TARGET(codePtr)
+                           : "rdx", "r8", "r9", "r10", "r11", "cc", "memory");
+#else
      __asm__ __volatile__("call *%%rax"
                           : "=a" (raxGetsWiped),
                             "=c" (rcxGetsWiped)
                           : "0" (codePtr),
                             "1" (crosspage)
                           : "rdx", "r8", "r9", "r10", "r11", "cc", "memory");
+#endif
+
   }
 #elif defined(_MSC_VER)
   /*
@@ -453,7 +453,9 @@ typedef enum {
 #define MSR_K7_HWCR_SSEDIS         0x00008000ULL // Disable SSE bit
 #define MSR_K7_HWCR_MONMWAITUSEREN 0x00000400ULL // Enable MONITOR/MWAIT CPL>0
 #define MSR_K7_HWCR_TLBFFDIS       0x00000040ULL // Disable TLB Flush Filter
+#ifndef MSR_K7_HWCR_SMMLOCK
 #define MSR_K7_HWCR_SMMLOCK        0x00000001ULL // Lock SMM environment
+#endif

 #ifndef MSR_K8_SYSCFG
 #define MSR_K8_SYSCFG        0xc0010010
@@ -981,7 +981,7 @@ LinuxDriverReadTSC(void *data,   // OUT: TSC values
 *-----------------------------------------------------------------------------
 */

-__attribute__((always_inline)) static Bool
+__always_inline static Bool
 LinuxDriverSyncReadTSCs(uint64 *delta) // OUT: TSC max - TSC min
 {
   TSCDelta tscDelta;
@@ -32,6 +32,6 @@ $(DRIVER)-y := driver.o hub.o userif.o netif.o bridge.o procfs.o smac_compat.o \
 ####

 clean:
-	rm -rf $(DRIVER).o $(DRIVER).mod.o $(DRIVER).mod.c $(DRIVER).ko \
+	rm -rf $(DRIVER).o $(DRIVER).mod.o $(DRIVER).mod.c $(DRIVER).ko .cache.mk \
 	       .tmp_versions Module.symvers Modules.symvers Module.markers \
 	       modules.order $($(DRIVER)-y) .*.cmd .*.o.flags
Author	SHA1	Message	Date
Michal Kubecek	db066be09b	vmmon: fix indirect call with retpoline build Build against kernel with retpoline support issues warning objtool: Task_Switch()+0x425: indirect call found in RETPOLINE build This is because an indirect call in TaskSwitchToMonitor() is encoded using inline assembler so that it bypasses retpoline generation. For this purpose, macro CALL_NOSPEC exists since v4.15-rc8 (and has been backported into some distribution kernels with the rest of retpoline support). Use the macro if available and fallback to the original code if not.	2018-08-20 07:57:39 +02:00
Michal Kubecek	036bbbe820	vmmon: fix always_inline attribute usage Function declared with __attribute__((always_inline)) should also be declared as inline, otherwise gcc issues a warning "always_inline function might not be inlinable". It's just cosmetic but getting rid of known harmless warnings makes it easier to spot actual problems. Use the __always_inline macro for LinuxDriverSyncReadTSCs() as this is how always inline functions should be declared in kernel code.	2018-08-20 07:57:39 +02:00
Michal Kubecek	18254c1bdf	vmmon: use standard definition of MSR_K7_HWCR_SMMLOCK if available The MSR_K7_HWCR_SMMLOCK macro is defined in mainline since commit 18c71ce9c882 ("x86/CPU/AMD: Add the Secure Encrypted Virtualization CPU feature") in v4.16-rc1.	2018-08-20 07:57:39 +02:00
Theo Bertacchini	457ae29811	modules: remove .cache.mk on make clean Remove .cache.mk inside vmmon-only and vmnet-only when executing "make clean". This file can cause issues when upgrading gcc as make will still look for includes inside older gcc includes directory. File .cache.mk with cache of generated variables is created by build since kernel v4.15-rc1, commit 3298b690b21c ("kbuild: Add a cache for generated variables").	2018-08-20 07:57:39 +02:00
Michal Kubecek	4dc79c5e97	vmmon: quick workaround for objtool warnings As discussed in https://bugzilla.suse.com/show_bug.cgi?id=1059674 the reason for multiple objtool warnings is the fact that vmmon module defines its own Panic() function which never returns. While it is marked as such which is used by the compiler for optimization, there is no way to find this from object file. While this seems innocuous, it might result in problems with unwinder later. The quickest way around is to replace vmmon's own Panic() with standard kernel panic() until a cleaner solution is found.	2018-08-20 07:57:39 +02:00
Michal Kubecek	1d6e4154c3	vmmon: use standard definition of CR3_PCID_MASK if available The CR3_PCID_MASK is defined in mainline asm/processor-flags.h since commit 6c690ee1039b ("x86/mm: Split read_cr3() into read_cr3_pa() and __read_cr3()") in v4.13-rc1.	2018-08-20 07:57:39 +02:00
Michal Kubecek	c27c9adc99	vmmon: use standard definition of MSR_MISC_FEATURES_ENABLES if available The MSR_MISC_FEATURES_ENABLES macro is defined in mainline since commit ab6d9468631a ("x86/msr: Rename MISC_FEATURE_ENABLES to MISC_FEATURES_ENABLES") in v4.12-rc1. The MSR_MISC_FEATURES_ENABLES_CPUID_FAULT macro is defined in mainline since commit e9ea1e7f53b8 ("x86/arch_prctl: Add ARCH_[GET\|SET]_CPUID") in v4.12-rc1.	2018-08-20 07:57:39 +02:00
Michal Kubecek	8416dacd5e	vmnet: use standard definition of PCI_VENDOR_ID_VMWARE_VMXNET3 if available The PCI_DEVICE_ID_VMWARE_VMXNET3 macro is defined in mainline pci_ids.h since commit b1226c7db1d9 ("vmxnet3: Move PCI Id to pci_ids.h") in v4.10-rc1.	2018-08-20 07:57:39 +02:00
Michal Kubecek	5c073f9eda	vmnet: use standard definition of PCI_VENDOR_ID_VMWARE if available The PCI_VENDOR_ID_VMWARE macro is defined in mainline pci_ids.h since commit 94e57fea6202 ("PCI: Move PCI_VENDOR_ID_VMWARE to pci_ids.h") in v3.18-rc1.	2018-08-20 07:57:39 +02:00
Michal Kubecek	3f188d0a76	import Workstation 14.1.3 module sources	2018-08-20 07:30:12 +02:00
Michal Kubecek	454cbd7b96	Merge branch 'master' into workstation	2018-05-27 16:13:45 +02:00
Michal Kubecek	2749950c5f	import Workstation 14.1.2 module sources	2018-05-23 18:49:52 +02:00
Michal Kubecek	7fa9cbe769	Merge branch 'master' into workstation	2018-04-02 15:37:06 +02:00
Michal Kubecek	cb66161bb5	Merge branch 'master' into workstation	2018-02-04 21:50:23 +01:00
Michal Kubecek	876363be72	import Workstation 14.1.1 module sources	2018-01-12 13:51:07 +01:00
Michal Kubecek	f78dd5288e	import Workstation 14.1.0 module sources	2017-12-25 10:06:21 +01:00
Michal Kubecek	e4df7e3461	import Workstation 14.0.0 module sources	2017-09-30 21:05:11 +02:00
Michal Kubecek	2bd541e265	import Workstation 12.5.7 module sources	2017-06-23 09:12:56 +02:00
Michal Kubecek	2a214307b9	import Workstation 12.5.6 module sources	2017-05-31 10:17:49 +02:00
Michal Kubecek	755b521dd1	import Workstation 12.5.5 module sources	2017-05-31 09:36:09 +02:00