- 18 Jun, 2021 4 commits
-
-
Heiko Carstens authored
s390 is the only architecture which makes use of the __no_kasan_or_inline attribute for two functions. Given that both stap() and __load_psw_mask() are very small functions they can and should be always inlined anyway. Therefore get rid of __no_kasan_or_inline and always inline these functions. Signed-off-by:
Heiko Carstens <hca@linux.ibm.com> Signed-off-by:
Vasily Gorbik <gor@linux.ibm.com>
-
Peter Oberparleiter authored
When read via debugfs, s390dbf debug-views print the kernel address of the call-site that created a trace entry. The kernel's %p pointer hashing feature obfuscates this address, and commit 860ec7c6 ("s390/debug: use pK for kernel pointers") made this obfuscation configurable via the kptr_restrict sysctl. Obfuscation of kernel address data printed via s390dbf debug-views does not add any additional protection since the associated debugfs files are only accessible to the root user that typically has enough other means to obtain kernel address data. Also trace payload data may contain binary representations of kernel addresses as part of logged data structues. Requiring such payload data to be obfuscated as well would be impractical and greatly diminish the use of s390dbf. Therefore completely remove pointer obfuscation from s390dbf debug-views. Reviewed-by:
Steffen Maier <maier@linux.ibm.com> Acked-by:
Peter Oberparleiter <oberpar@linux.ibm.com> Signed-off-by:
-
Vasily Gorbik authored
Since OLDMEM_BASE/OLDMEM_SIZE is already taken into consideration and is reflected in ident_map_size. reserve/remove_oldmem() is no longer needed and could be removed. Reviewed-by:
Alexander Gordeev <agordeev@linux.ibm.com> Signed-off-by:
-
Vasily Gorbik authored
Currently there are two separate places where kernel memory layout has to be known and adjusted: 1. early kasan setup. 2. paging setup later. Those 2 places had to be kept in sync and adjusted to reflect peculiar technical details of one another. With additional factors which influence kernel memory layout like ultravisor secure storage limit, complexity of keeping two things in sync grew up even more. Besides that if we look forward towards creating identity mapping and enabling DAT before jumping into uncompressed kernel - that would also require full knowledge of and control over kernel memory layout. So, de-duplicate and move kernel memory layout setup logic into the decompressor. Reviewed-by:
-
- 16 Jun, 2021 2 commits
-
-
Alexander Gordeev authored
Introduce OUTSIDE macro that checks whether an instruction address is inside or outside of a block of instructions. Signed-off-by:
-
Harald Freudenberger authored
This patch brings 3 reworked/new uevent changes: * All AP uevents caused by an ap card or queue device now carry an additional uevent env value MODE=<accel|cca|ep11>. Here is an example: KERNEL[1267.301292] add /devices/ap/card0a (ap) ACTION=add DEVPATH=/devices/ap/card0a SUBSYSTEM=ap DEVTYPE=ap_card DEV_TYPE=000D MODALIAS=ap:t0D MODE=ep11 <- this is new SEQNUM=1095 This is true for bind, unbind, add, remove, and change uevents related to ap card or ap queue devices. * On a change of the soft online attribute on a zcrypt queue or card device a new CHANGE uevent is sent with an env value ONLINE=<0|1>. Example uevent: KERNEL[613.067531] change /devices/ap/card09/09.0011 (ap) ACTION=change DEVPATH=/devices/ap/card09/09.0011 SUBSYSTEM=ap ONLINE=0 <- this is new DEVTYPE=ap_queue DRIVER=cex4queue MODE=cca SEQNUM=1070 - On a change of the config state of an zcrypt card device a new CHANGE uevent is sent with an env value CONFIG=<0|1>. Example uevent: KERNEL[876.258680] change /devices/ap/card09 (ap) ACTION=change DEVPATH=/devices/ap/card09 SUBSYSTEM=ap CONFIG=0 <- this is new DEVTYPE=ap_card DRIVER=cex4card DEV_TYPE=000D MODALIAS=ap:t0D MODE=cca SEQNUM=1073 Setting a card config on/off causes the dependent queue devices to follow the config state change and thus uevents informing about the config state change for the queue devices are also emitted. Signed-off-by:Harald Freudenberger <freude@linux.ibm.com> Reviewed-by:
Ingo Franzki <ifranzki@linux.ibm.com> Signed-off-by:
-
- 07 Jun, 2021 33 commits
-
-
Rob Landley authored
Use simple sed/tr instead of perl to generate decompressor symbols file with the same result. Signed-off-by:
Rob Landley <rob@landley.net> Signed-off-by:
-
Alexander Gordeev authored
cleanup_sie_mcck label is called from a single location only and thus does not need to be a subroutine. Move the labelled code to the caller - by doing that the SIE critical section checks appear next to each other and the SIE cleanup becomes bit more readable. Signed-off-by:
-
Vasily Gorbik authored
This helps to avoid merge conflicts later. * fixes: s390/mcck: fix invalid KVM guest condition check s390/mcck: fix calculation of SIE critical section size Signed-off-by: -
Heiko Carstens authored
Signed-off-by:
-
Sven Schnelle authored
Signed-off-by:
Sven Schnelle <svens@linux.ibm.com> Signed-off-by:
-
Sven Schnelle authored
To avoid casting ptrace members, add a union containing both struct tpi_info and explicit int_code/int_parm members. Signed-off-by:
-
Sven Schnelle authored
For future work we need the struct tpi_info declaration in asm/ptrace.h. Due to circular dependencies it cannot stay in asm/lowcore.h or asm/cio.h, which would be the preferred location. Therefore add it in its own header file. Also fix a typo in the length of a reserved field that did not have a functional effect beyond an incorrect field value in the s390_cio_tpi tracepoint. Fixes: 2ab59de7 ("s390/cio: Consolidate inline assemblies and related data definitions") Signed-off-by:
-
Randy Dunlap authored
APPLDATA_BASE should depend on PROC_SYSCTL instead of PROC_FS. Building with PROC_FS but not PROC_SYSCTL causes a build error, since appldata_base.c uses data and APIs from fs/proc/proc_sysctl.c. arch/s390/appldata/appldata_base.o: in function `appldata_generic_handler': appldata_base.c:(.text+0x192): undefined reference to `sysctl_vals' Fixes: c185b783 ("[S390] Remove config options.") Signed-off-by:
Randy Dunlap <rdunlap@infradead.org> Cc: Heiko Carstens <hca@linux.ibm.com> Cc: Vasily Gorbik <gor@linux.ibm.com> Cc: Christian Borntraeger <borntraeger@de.ibm.com> Cc: linux-s390@vger.kernel.org Signed-off-by:
-
Fabrice Fontaine authored
Though -nostdlib is passed in PURGATORY_LDFLAGS and -ffreestanding in KBUILD_CFLAGS_DECOMPRESSOR, -fno-stack-protector must also be passed to avoid linking errors related to undefined references to '__stack_chk_guard' and '__stack_chk_fail' if toolchain enforces -fstack-protector. Fixes - https://gitlab.com/kubu93/buildroot/-/jobs/1247043361Signed-off-by:
Fabrice Fontaine <fontaine.fabrice@gmail.com> Signed-off-by:
Alexander Egorenkov <egorenar@linux.ibm.com> Tested-by:
-
Julian Wiedmann authored
ccwgroup_notifier() currently listens for BUS_NOTIFY_UNBIND_DRIVER events, and triggers an ungrouping for the affected device. Looking at __device_release_driver(), we can wait for a little longer until the driver has been fully unbound and eg. bus->remove() has been called. So listen for BUS_NOTIFY_UNBOUND_DRIVER instead. Due to locking the current code should work just fine, but this clarifies our intent. Signed-off-by:
Julian Wiedmann <jwi@linux.ibm.com> Reviewed-by:
Vineeth Vijayan <vneethv@linux.ibm.com> Signed-off-by:
-
Julian Wiedmann authored
driver_unregister() ends up calling __device_release_driver() for each device that is bound to the driver. Thus ccwgroup_notifier() will see a BUS_NOTIFY_UNBIND_DRIVER event for these ccwgroup devices, and trigger the ungrouping. So there's no need to also trigger the ungrouping from within ccwgroup_driver_unregister(), remove it. Signed-off-by:
-
Julian Wiedmann authored
Define & initialize the driver struct in one go, so that everything is in one place. Signed-off-by:
Halil Pasic <pasic@linux.ibm.com> Signed-off-by:
-
Harald Freudenberger authored
Userspace udev rules need an indication about the very first AP change BINDINGS=complete uevent. So now this uevent is extend with an additional key-value entry COMPLETECOUNT=<counter>. The very first uevent will show counter=1 and the following BINDINGS=complete uevents increase this counter by 1. Here is an example how the very first BINDINGS=complete uevent looks like: KERNEL[106.079510] change /devices/ap (ap) ACTION=change DEVPATH=/devices/ap SUBSYSTEM=ap BINDINGS=complete COMPLETECOUNT=1 SEQNUM=10686 Signed-off-by:
-
Alexander Gordeev authored
Per-CPU pointer to lowcore is stored in global lowcore_ptr[] array and duplicated in struct pcpu::lowcore member. This update removes the redundancy. Signed-off-by:
-
Alexander Gordeev authored
Once the kernel is running the boot CPU lowcore becomes freeable and does not differ from the secondary CPU ones in any way. Make use of it and do not preserve the boot CPU lowcore on unplugging. That allows returning unused memory when the boot CPU is offline and makes the code more clear. Signed-off-by:
-
Alexander Gordeev authored
The lowcore for IPL CPU is special. It is allocated early in the boot process using memblock and never freed since. The reason is pcpu_alloc_lowcore() and pcpu_free_lowcore() routines use page allocator which is not available when the IPL CPU is getting initialized. Similar problem is already addressed for stacks - once the virtual memory is available the early boot stacks get re- allocated. Doing the same for lowcore will allow freeing the IPL CPU lowcore and make no difference between the boot and secondary CPUs. Signed-off-by:
-
Alexander Gordeev authored
Architecture callback switch_mm() is allowed to be called with enabled interrupts. However, our implementation of switch_mm() does not expect that. Let's follow other architectures and make sure switch_mm() is always executed with interrupts disabled, regardless of what happens with the generic kernel code in the future. Signed-off-by:
-
Heiko Carstens authored
Access the kernel command line via parmarea instead of using the COMMAND_LINE define. With this the following gcc11 warning is removed: arch/s390/boot/ipl_parm.c: In function ‘setup_boot_command_line’: arch/s390/boot/ipl_parm.c:168:50: warning: writing 1 byte into a region of size 0 [-Wstringop-overflow=] Signed-off-by:
-
Heiko Carstens authored
Since commit 9a965ea95135 ("s390/kexec_file: Simplify parmarea access") we have struct parmarea which describes the layout of the kernel parameter area. Make the kernel parameter area available as global variable parmarea of type struct parmarea, which allows to easily access its members. Signed-off-by: -
Valentin Vidic authored
Console name reported in /proc/consoles: ttyS1 -W- (EC p ) 4:65 does not match the char device name: crw--w---- 1 root root 4, 65 May 17 12:18 /dev/ttysclp0 so debian-installer inside a QEMU s390x instance gets confused and fails to start with the following error: steal-ctty: No such file or directory Signed-off-by:
Valentin Vidic <vvidic@valentin-vidic.from.hr> Link: https://lore.kernel.org/r/20210427194010.9330-1-vvidic@valentin-vidic.from.hrSigned-off-by:
Christian Borntraeger <borntraeger@de.ibm.com> Signed-off-by:
-
Julian Wiedmann authored
Wiring up the cdevs is an essential part of the gdev creation. So release them during the gdev destruction, ie. on the last put_device(). This could cause us to hold on to the cdevs a tiny bit longer, but that's not a real concern. As we have much less certainty of what context this put_device() is executed from, switch to irqsave locking. Signed-off-by:
-
Julian Wiedmann authored
Hijacking the device_driver's probe/remove callbacks for purely bus-internal logic is a very unconvential construct. Instead just set up our callbacks in the AP bus struct, and really_probe() will call them in the same way as before. Signed-off-by:
-
Julian Wiedmann authored
It's evidently unused. Signed-off-by:
-
Sven Schnelle authored
gcc-11 warns: drivers/s390/crypto/zcrypt_ccamisc.c:298:38: warning: argument 4 of type u8[64] {aka unsigned char[64]} with mismatched bound [-Warray-parameter=] 298 | u32 keybitsize, u8 seckey[SECKEYBLOBSIZE]) | ~~~^~~~~~~~~~~~~~~~~~~~~~ In file included from drivers/s390/crypto/zcrypt_ccamisc.c:24: drivers/s390/crypto/zcrypt_ccamisc.h:162:63: note: previously declared as u8 * {aka unsigned char *} 162 | int cca_genseckey(u16 cardnr, u16 domain, u32 keybitsize, u8 *seckey); | ~~~~^~~~~~ drivers/s390/crypto/zcrypt_ccamisc.c:441:41: warning: argument 5 of type u8[64] {aka unsigned char[64]} with mismatched bound [-Warray-parameter=] 441 | const u8 *clrkey, u8 seckey[SECKEYBLOBSIZE]) | ~~~^~~~~~~~~~~~~~~~~~~~~~ In file included from drivers/s390/crypto/zcrypt_ccamisc.c:24: drivers/s390/crypto/zcrypt_ccamisc.h:168:42: note: previously declared as u8 * {aka unsigned char *} 168 | const u8 *clrkey, u8 *seckey); | ~~~~^~~~~~ Fix this by making the prototypes match the functions. Signed-off-by: -
Sven Schnelle authored
gcc-11 warns: arch/s390/kernel/traps.c: In function __do_pgm_check: arch/s390/kernel/traps.c:319:17: warning: memcpy reading 256 bytes from a region of size 0 [-Wstringop-overread] 319 | memcpy(¤t->thread.trap_tdb, &S390_lowcore.pgm_tdb, 256); | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Fix this by adding a struct pgm_tdb to struct lowcore and copy that. Signed-off-by: -
Sven Schnelle authored
gcc-11 warns: arch/s390/kernel/irq.c: In function do_ext_irq: arch/s390/kernel/irq.c:175:9: warning: memcpy reading 4 bytes from a region of size 0 [-Wstringop-overread] 175 | memcpy(®s->int_code, &S390_lowcore.ext_cpu_addr, 4); | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Fix this by adding a struct for int_code to struct lowcore. Signed-off-by: -
Sven Schnelle authored
With gcc-11, there are a lot of warnings because the facility functions are accessing lowcore through a null pointer. Fix this by moving the facility arrays away from lowcore. Signed-off-by:
-
Sven Schnelle authored
arch/s390/kernel/syscall.c: In function __do_syscall: arch/s390/kernel/syscall.c:147:9: warning: memcpy reading 64 bytes from a region of size 0 [-Wstringop-overread] 147 | memcpy(®s->gprs[8], S390_lowcore.save_area_sync, 8 * sizeof(unsigned long)); | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ arch/s390/kernel/syscall.c:148:9: warning: memcpy reading 4 bytes from a region of size 0 [-Wstringop-overread] 148 | memcpy(®s->int_code, &S390_lowcore.svc_ilc, sizeof(regs->int_code)); | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Fix this by moving the gprs restore from C to assembly, and use a assignment for int_code instead of memcpy. Signed-off-by: -
Niklas Schnelle authored
In commit b02002cc ("s390/pci: Implement ioremap_wc/prot() with MIO") we implemented both ioremap_wc() and ioremap_prot() however until now we had not set HAVE_IOREMAP_PROT in Kconfig, do so now. This also requires implementing pte_pgprot() as this is used in the generic_access_phys() code enabled by CONFIG_HAVE_IOREMAP_PROT. As with ioremap_wc() we need to take the MMIO Write Back bit index into account. Moreover since the pgprot value returned from pte_pgprot() is to be used for mappings into kernel address space we must make sure that it uses appropriate kernel page table protection bits. In particular a pgprot value originally coming from userspace could have the _PAGE_PROTECT bit set to enable fault based dirty bit accounting which would then make the mapping inaccessible when used in kernel address space. Fixes: b02002cc ("s390/pci: Implement ioremap_wc/prot() with MIO") Reviewed-by:
Gerald Schaefer <gerald.schaefer@linux.ibm.com> Signed-off-by:
Niklas Schnelle <schnelle@linux.ibm.com> Signed-off-by:
-
Thomas Richter authored
Remove some WARN_ON_ONCE() warnings when a counter is started. Each counter is installed function calls event_sched_in() --> cpumf_pmu_add(..., PERF_EF_START). This is done after the event has been created using perf_pmu_event_init() which verifies the counter is valid. Member hwc->config must be valid at this point. Function cpumf_pmu_start(..., PERF_EF_RELOAD) is called from function cpumf_pmu_add() for counter events. All other invocations of cpumf_pmu_start(..., PERF_EF_RELOAD) are from the performance subsystem for sampling events. Signed-off-by:
Thomas Richter <tmricht@linux.ibm.com> Acked-by:
Sumanth Korikkar <sumanthk@linux.ibm.com> Signed-off-by:
-
Thomas Richter authored
The command 'perf stat -e cycles ...' triggers the following function sequence in the CPU Measurement Facility counter device driver: perf_pmu_event_init() __hw_perf_event_init() validate_ctr_auth() validate_ctr_version() During event creation, the counter number is checked in functions validate_ctr_auth() and validate_ctr_version() to verify it is a valid counter and supported by the hardware. If this is not the case, both functions return an error and the event is not created. System call perf_event_open() returns an error in this case. Later on the event is installed in the kernel event subsystem and the driver functions cpumf_pmu_add() and cpumf_pmu_commit_txn() are called to install the counter event by the hardware. Since both events have been verified at event creation, there is no need to re-evaluate the authorization state. This can not change since on * LPARs the authorization change requires a restart of the LPAR (and thus a reboot of the kernel) * DPMs can not take resources away, just add them. Also the sequence of CPU Measurement facility counter device driver calls is cpumf_pmu_start_txn cpumf_pmu_add cpumf_pmu_start cpumf_pmu_commit_txn for every single event. Which means the condition in cpumf_pmu_add() is never met and validate_ctr_auth() is never called. This leaves the counter device driver transaction functions with just one task: start_txn: Verify a transaction is not in flight and call perf_pmu_disable() cancel_txn, commit_txn: Verify a transaction is in flight and call perf_pmu_enable() The same functionality is provided by the default transaction handling functions in kernel/events/core.c. Use those by removing the counter device driver private call back functions. Suggested-by: -
Alexander Gordeev authored
Wrong condition check is used to decide if a machine check hit while in KVM guest. As result of this check the instruction following the SIE critical section might be considered as still in KVM guest and _CIF_MCCK_GUEST CPU flag mistakenly set as result. Fixes: c929500d ("s390/nmi: s390: New low level handling for machine check happening in guest") Cc: <stable@vger.kernel.org> Signed-off-by:
-
Alexander Gordeev authored
The size of SIE critical section is calculated wrongly as result of a missed subtraction in commit 0b0ed657 ("s390: remove critical section cleanup from entry.S") Fixes: 0b0ed657 ("s390: remove critical section cleanup from entry.S") Cc: <stable@vger.kernel.org> Signed-off-by:
-
- 06 Jun, 2021 1 commit
-
-
Linus Torvalds authored
-
