- 24 Nov, 2022 40 commits
-
-
Tiezhu Yang authored
The latest version of grep claims the egrep is now obsolete so the build now contains warnings that look like: egrep: warning: egrep is obsolescent; using grep -E fix this up by moving the related file to use "grep -E" instead. sed -i "s/egrep/grep -E/g" `grep egrep -rwl arch/powerpc` Here are the steps to install the latest grep: wget http://ftp.gnu.org/gnu/grep/grep-3.8.tar.gz tar xf grep-3.8.tar.gz cd grep-3.8 && ./configure && make sudo make install export PATH=/usr/local/bin:$PATH Signed-off-by:
Tiezhu Yang <yangtiezhu@loongson.cn> Signed-off-by:
Michael Ellerman <mpe@ellerman.id.au> Link: https://lore.kernel.org/r/1668764429-11540-1-git-send-email-yangtiezhu@loongson.cn
-
Nayna Jain authored
Even though plpks_read_var() is currently called to read variables owned by different consumers, it internally supports only OS consumer. Fix plpks_read_var() to handle different consumers correctly. Fixes: 2454a7af ("powerpc/pseries: define driver for Platform KeyStore") Signed-off-by:
Nayna Jain <nayna@linux.ibm.com> Signed-off-by:
-
Nayna Jain authored
Replace kmalloc with kzalloc in construct_auth() function to default initialize structure with zeroes. Signed-off-by:
-
Nayna Jain authored
Logging H_CALL return codes in PLPKS driver are easy to confuse with Linux error codes. Let the caller of the function log the converted linux error code. Signed-off-by:
-
Nayna Jain authored
Some commands for eg. "cat" might continue to retry on encountering EINTR. This is not expected for original error code H_ABORTED. Map H_ABORTED to more relevant Linux error code EIO. Fixes: 2454a7af ("powerpc/pseries: define driver for Platform KeyStore") Signed-off-by:
-
Nayna Jain authored
PAPR Spec defines H_P1 actually as H_PARAMETER and maps H_ABORTED to a different numerical value. Fix the error codes as per PAPR Specification. Fixes: 2454a7af ("powerpc/pseries: define driver for Platform KeyStore") Signed-off-by:
-
Nayna Jain authored
OS_VAR_LINUX enum in PLPKS driver should be 0x02 instead of 0x01. Fixes: 2454a7af ("powerpc/pseries: define driver for Platform KeyStore") Signed-off-by:
-
Chen Lifu authored
1. ppc_override_l2cr and ppc_override_l2cr_value are only used in l2cr_init() function, remove them and used *l2cr directly. 2. has_l2cache is not used outside of the file, so mark it static and do not initialise statics to 0. Fixes the following warnings: arch/powerpc/platforms/powermac/setup.c:73:5: warning: symbol 'ppc_override_l2cr' was not declared. Should it be static? arch/powerpc/platforms/powermac/setup.c:74:5: warning: symbol 'ppc_override_l2cr_value' was not declared. Should it be static? arch/powerpc/platforms/powermac/setup.c:75:5: warning: symbol 'has_l2cache' was not declared. Should it be static? Signed-off-by:
Chen Lifu <chenlifu@huawei.com> Reviewed-by:
Christophe Leroy <christophe.leroy@csgroup.eu> [mpe: Unwrap printk string] Signed-off-by:
-
Bo Liu authored
Fixes the following W=1 kernel build warning(s): arch/powerpc/platforms/pseries/eeh_pseries.c:163: warning: Function parameter or member 'config_addr' not described in 'pseries_eeh_phb_reset' arch/powerpc/platforms/pseries/eeh_pseries.c:163: warning: Excess function parameter 'config_adddr' description in 'pseries_eeh_phb_reset' arch/powerpc/platforms/pseries/eeh_pseries.c:198: warning: Function parameter or member 'config_addr' not described in 'pseries_eeh_phb_configure_bridge' arch/powerpc/platforms/pseries/eeh_pseries.c:198: warning: Excess function parameter 'config_adddr' description in 'pseries_eeh_phb_configure_bridge' Signed-off-by:
Bo Liu <liubo03@inspur.com> Signed-off-by:
-
Russell Currey authored
In hw_breakpoint_handler(), ea is set by wp_get_instr_detail() except for 8xx, leading the variable to be passed uninitialised to wp_check_constraints(). This is safe as wp_check_constraints() returns early without using ea, so just set it to make the compiler happy. Signed-off-by:
Russell Currey <ruscur@russell.cc> Reviewed-by:
-
Shaomin Deng authored
Remove the repeated word "not" in comments. Signed-off-by:
Shaomin Deng <dengshaomin@cdjrlc.com> Signed-off-by:
-
Colin Ian King authored
There is a spelling mistake in a perror message. Fix it. Signed-off-by:
Colin Ian King <colin.i.king@gmail.com> Signed-off-by:
-
Naveen N. Rao authored
preempt_enable_no_resched() is just the same as preempt_enable() when we are in a irqs disabled context. kprobe_handler() and the post/fault handlers are all called with irqs disabled. As such, convert those to just use preempt_enable(). Reported-by:
Nicholas Piggin <npiggin@gmail.com> Signed-off-by:
Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Reviewed-by:
-
Naveen N. Rao authored
Similar to x86 commit 2e62024c ("kprobes/x86: Use preempt_enable() in optimized_callback()"), change powerpc optprobes to use preempt_enable() rather than preempt_enable_no_resched() since powerpc also removed irq disabling for optprobes in commit f72180cc ("powerpc/kprobes: Do not disable interrupts for optprobes and kprobes_on_ftrace"). Reported-by:
-
Naveen N. Rao authored
arch_prepare_kprobe() is called from register_kprobe() via prepare_kprobe(), or through register_aggr_kprobe(), both with the kprobe_mutex held. Per the comment for get_kprobe(): /* * This routine is called either: * - under the 'kprobe_mutex' - during kprobe_[un]register(). * OR * - with preemption disabled - from architecture specific code. */ As such, there is no need to disable preemption around the call to get_kprobe(). Drop the same. Reported-by:
-
ruanjinjie authored
When build Linux kernel, encounter the following warnings: ./arch/powerpc/sysdev/mpic_msgr.c:230:38: warning: cast removes address space '__iomem' of expression ./arch/powerpc/sysdev/mpic_msgr.c:230:27: warning: incorrect type in assignment (different address spaces) The data type of msgr->mer and msgr->base are 'u32 __iomem *', but converted to 'u32 *' and 'u8 *' directly and cause above warnings, now instead of using a type cast, change the size of the pointer offset to fix these warnings. Signed-off-by:
ruanjinjie <ruanjinjie@huawei.com> Reviewed-by:
-
Yang Yingliang authored
If remapping 'data->trig_page' fails, the 'data->eoi_mmio' need be unmapped before returning from xive_spapr_populate_irq_data(). Fixes: eac1e731 ("powerpc/xive: guest exploitation of the XIVE interrupt controller") Signed-off-by:
Yang Yingliang <yangyingliang@huawei.com> Reviewed-by:
Cédric Le Goater <clg@kaod.org> Signed-off-by:
-
Stephen Rothwell authored
This is a follow on from commit 0d362be5 ("Makefile: link with -z noexecstack --no-warn-rwx-segments") for arch/powerpc/boot to address wanrings like: ld: warning: opal-calls.o: missing .note.GNU-stack section implies executable stack ld: NOTE: This behaviour is deprecated and will be removed in a future version of the linker ld: warning: arch/powerpc/boot/zImage.epapr has a LOAD segment with RWX permissions This fixes issue https://github.com/linuxppc/issues/issues/417Signed-off-by:
Stephen Rothwell <sfr@canb.auug.org.au> Signed-off-by:
-
Christophe JAILLET authored
At the beginning of the file, we have: #define pr_fmt(fmt) "xive: " fmt So, there is no need to duplicate "XIVE:" in debug and error messages. For the records, these useless prefix have been added in commit 5af50993 ("KVM: PPC: Book3S HV: Native usage of the XIVE interrupt controller") Signed-off-by:
Christophe JAILLET <christophe.jaillet@wanadoo.fr> Signed-off-by:
-
Nicholas Piggin authored
cputime_t was a core kernel type, removed by commits ed5c8c85..b672592f. As explained in commit b672592f ("sched/cputime: Remove generic asm headers"), the final cleanup is for the arch to provide cputime_to_nsec[s](). Commit ade7667a ("powerpc: Add cputime_to_nsecs()") did that, but justdidn't remove the then-unused cputime_to_usecs(), cputime_t type, and associated remnants. Signed-off-by:
-
Michael Ellerman authored
Although the previous commit made the powerpc instruction dump usable with scripts/decodecode, there are still some problems. Because the dump is split across multiple lines, the script doesn't cope with printk timestamps or caller info. That can be fixed by printing the entire dump on one line, eg: [ 12.016307][ T112] --- interrupt: c00 [ 12.016605][ T112] Code: 4b7aae15 60000000 3d22016e 3c62ffec 39291160 38639bc0 e8890000 4b7aadf9 60000000 4bfffee8 7c0802a6 60000000 <0fe00000> 60420000 3c4c008f 384268a0 [ 12.017655][ T112] ---[ end trace 0000000000000000 ]--- That output can then be piped directly into scripts/decodecode and interpreted correctly. Printing the dump on a single line does produce a very long line, about 173 characters. That is still shorter than x86, which prints nearly 200 characters even without timestamps etc. All consoles I'm aware of will wrap the line if it's too long, so the length should not be a functional problem. If anything it should help on consoles like VGA by using less vertical space. Signed-off-by:
-
Michael Ellerman authored
Matt reported that scripts/decodecode doesn't work for the instruction dump in the powerpc oops output. Although there are scripts around that can decode it, it would be preferable if the standard in-tree script worked. All other arches prefix the instruction dump with "Code:", and that's what the script looks for, so use that. The script then works as expected: $ CROSS_COMPILE=powerpc64le-linux-gnu- ./scripts/decodecode Code: fbc1fff0 f821ffc1 7c7d1b78 7c9c2378 ebc30028 7fdff378 48000018 60000000 60000000 ebff0008 7c3ef840 41820048 <815f0060> e93f0000 5529077c 7d295378 ^D All code ======== 0: f0 ff c1 fb std r30,-16(r1) 4: c1 ff 21 f8 stdu r1,-64(r1) 8: 78 1b 7d 7c mr r29,r3 ... Note that the script doesn't cope well with printk timestamps or printk caller info. Reported-by:Matthew Wilcox <willy@infradead.org> Signed-off-by:
-
Joel Stanley authored
This is the register layout of the litesd peripheral for the fusesoc based Microwatt SoC. It requires a description of the system clock, which is hardcoded to 100MHz. Signed-off-by:
Joel Stanley <joel@jms.id.au> Signed-off-by:
-
Christophe Leroy authored
At the time being, with 16k pages __set_pte_at() writes table entries in reverse order: 294: 91 49 00 0c stw r10,12(r9) 298: 91 49 00 08 stw r10,8(r9) 29c: 91 49 00 04 stw r10,4(r9) 2a0: 91 49 00 00 stw r10,0(r9) Allthough there should be no impact at all as it stays in a single cacheline, reverse the writing in a more natural order. 288: 91 49 00 0c stw r10,0(r9) 28c: 91 49 00 08 stw r10,4(r9) 290: 91 49 00 04 stw r10,8(r9) 294: 91 49 00 00 stw r10,12(r9) Signed-off-by:
-
Christophe Leroy authored
While looking at code generated for code patching, I saw that pte_clear generated: 2d8: 38 a0 00 00 li r5,0 2dc: 38 e0 10 00 li r7,4096 2e0: 39 00 20 00 li r8,8192 2e4: 39 40 30 00 li r10,12288 2e8: 90 a9 00 00 stw r5,0(r9) 2ec: 90 e9 00 04 stw r7,4(r9) 2f0: 91 09 00 08 stw r8,8(r9) 2f4: 91 49 00 0c stw r10,12(r9) With 16k pages, only the first entry is used by the kernel, so no need to adapt the address of other entries. Only duplicate the first entry for hardware. Now it is: 2cc: 39 40 00 00 li r10,0 2d0: 91 49 00 00 stw r10,0(r9) 2d4: 91 49 00 04 stw r10,4(r9) 2d8: 91 49 00 08 stw r10,8(r9) 2dc: 91 49 00 0c stw r10,12(r9) Signed-off-by:
-
Dmitry Torokhov authored
This patch converts the driver to newer gpiod API, and away from OF-specific legacy gpio API that we want to stop using. While at it, let's address a few more issues: - switch to using dev_info()/pr_info() and friends - cancel work when unbinding the driver Note that the original code handled halt GPIO polarity incorrectly: in halt callback, when line polarity is "low" it would set trigger to "1" and drive halt line high, which is counter to the annotation. gpiod API will drive such line low. However I do not see any DTSes in mainline that have a DT node with "sgy,gpio-halt" compatible. Signed-off-by:
Dmitry Torokhov <dmitry.torokhov@gmail.com> Signed-off-by:
-
Dmitry Torokhov authored
This switches PIKA Warp away from legacy gpio API and to newer gpiod API, so that we can eventually deprecate the former. Because LEDs are normally driven by leds-gpio driver, but the platform code also wants to access the LEDs during thermal shutdown, and gpiod API does not allow locating GPIO without requesting it, the platform code is now responsible for locating GPIOs through device tree and requesting them. It then constructs platform data for leds-gpio platform device and registers it. This allows platform code to retain access to LED GPIO descriptors and use them when needed. Signed-off-by:
-
Gustavo A. R. Silva authored
When building with automatic stack variable initialization, GCC 12 complains about variables defined outside of switch case statements. Move the variable into the case that uses it, which silences the warning: arch/powerpc/xmon/xmon.c: In function ‘bpt_cmds’: arch/powerpc/xmon/xmon.c:1529:13: warning: statement will never be executed [-Wswitch-unreachable] 1529 | int mode; | ^~~~ Fixes: 09b6c112 ("powerpc/xmon: Fix compile error with PPC_8xx=y") Signed-off-by:Gustavo A. R. Silva <gustavoars@kernel.org> Reviewed-by:
Kees Cook <keescook@chromium.org> Signed-off-by:
-
Kajol Jain authored
Commit 57dc0eed ("KVM: PPC: Book3S HV P9: Implement PMU save/restore in C") removed the PMU save/restore functions from assembly code and implemented these functions in C, for power9 and later platforms. After the code refactoring, Performance Monitoring Unit (PMU) registers became part of "p9_host_os_sprs" structure and now this structure is used to save/restore pmu host registers, for power9 and later platfroms. But we still have old unused registers references. Patch removes unused host_mmcr references for Monitor Mode Control Register 3 (MMCR3)/ Sampled Instruction Event Register 2 (SIER2)/ SIER3 registers from "struct kvmppc_host_state". Fixes: 57dc0eed ("KVM: PPC: Book3S HV P9: Implement PMU save/restore in C") Signed-off-by:
Kajol Jain <kjain@linux.ibm.com> Reviewed-by:
Athira Rajeev <atrajeev@linux.vnet.ibm.com> Signed-off-by:
-
Disha Goel authored
The kvm code was refactored to convert some of kvm assembly routines to C. This includes commits which moved code path for the kvm guest entry/exit for p7/8 from aseembly to C. As part of the code changes, usage of some of the macros were removed. But definitions still exist in the assembly files. Commits are listed below: Commit 2e1ae9cd ("KVM: PPC: Book3S HV: Implement radix prefetch workaround by disabling MMU") Commit 9769a7fd ("KVM: PPC: Book3S HV: Remove radix guest support from P7/8 path") Commit fae5c9f3 ("KVM: PPC: Book3S HV: remove ISA v3.0 and v3.1 support from P7/8 path") Commit 57dc0eed ("KVM: PPC: Book3S HV P9: Implement PMU save/restore in C") Many of the asm-offset macro definitions were missed to remove. Patch fixes by removing the unused macros. Signed-off-by:
Disha Goel <disgoel@linux.vnet.ibm.com> Reviewed-by:
-
Xiu Jianfeng authored
Add missing __init/__exit annotations to module init/exit funcs. Signed-off-by:
Xiu Jianfeng <xiujianfeng@huawei.com> Signed-off-by:
-
Nicholas Piggin authored
ISA v2.06 (POWER7 and up) as well as e6500 support lbarx and lharx. Add a compile option that allows code to use it, and add support in cmpxchg and xchg 8 and 16 bit values without shifting and masking. Signed-off-by:
-
Shaomin Deng authored
Drop the repeated word "when" in comments. Signed-off-by:
-
Nicholas Miehlbradt authored
Add POWER9 and POWER10 to CPU families and list Radix MMU. Signed-off-by:
Nicholas Miehlbradt <nicholas@linux.ibm.com> Reviewed-by:
Andrew Donnellan <ajd@linux.ibm.com> Signed-off-by:
-
Naveen N. Rao authored
Extend commit b39181f7 ("ftrace: Add FTRACE_MCOUNT_MAX_OFFSET to avoid adding weak function") to ppc32 and ppc64 -mprofile-kernel by defining FTRACE_MCOUNT_MAX_OFFSET. For ppc64 -mprofile-kernel ABI, we can have two instructions at function entry for TOC setup followed by 'mflr r0' and 'bl _mcount'. So, the mcount location is at most the 4th instruction in a function. For ppc32, mcount location is always the 3rd instruction in a function, preceded by 'mflr r0' and 'stw r0,4(r1)'. With this patch, and with ppc64le_guest_defconfig and some ftrace/bpf config items enabled: # grep __ftrace_invalid_address available_filter_functions | wc -l 79 Signed-off-by:
Steven Rostedt (Google) <rostedt@goodmis.org> Signed-off-by:
-
Deming Wang authored
Operators should be separated by spaces in tce_buildmulti_pSeriesLP() Signed-off-by:
Deming Wang <wangdeming@inspur.com> Signed-off-by:
-
Yang Yingliang authored
get_function_0() calls pci_get_domain_bus_and_slot(), as comment says, it returns a pci device with refcount increment, so after using it, pci_dev_put() needs be called. Get the device reference when get_function_0() is not called, so pci_dev_put() can be called in the error path and callers unconditionally. And add comment above get_dvsec_vendor0() to tell callers to call pci_dev_put(). Fixes: 87db7579 ("ocxl: control via sysfs whether the FPGA is reloaded on a link reset") Suggested-by:
Frederic Barrat <fbarrat@linux.ibm.com> Acked-by:
-
Yang Yingliang authored
If device_register() returns error in ocxl_file_register_afu(), the name allocated by dev_set_name() need be freed. As comment of device_register() says, it should use put_device() to give up the reference in the error path. So fix this by calling put_device(), then the name can be freed in kobject_cleanup(), and info is freed in info_release(). Fixes: 75ca758a ("ocxl: Create a clear delineation between ocxl backend & frontend") Signed-off-by:
-
Yang Yingliang authored
If device_register() fails in cxl_pci_afu|adapter(), the device is not added, device_unregister() can not be called in the error path, otherwise it will cause a null-ptr-deref because of removing not added device. As comment of device_register() says, it should use put_device() to give up the reference in the error path. So split device_unregister() into device_del() and put_device(), then goes to put dev when register fails. Fixes: f204e0b8 ("cxl: Driver code for powernv PCIe based cards for userspace access") Signed-off-by:
-
Yang Yingliang authored
If device_register() fails in cxl_register_afu|adapter(), the device is not added, device_unregister() can not be called in the error path, otherwise it will cause a null-ptr-deref because of removing not added device. As comment of device_register() says, it should use put_device() to give up the reference in the error path. So split device_unregister() into device_del() and put_device(), then goes to put dev when register fails. Fixes: 14baf4d9 ("cxl: Add guest-specific code") Signed-off-by:
-
