From 6cf06c17e94f26c290fd3370a5c36514ae15ac43 Mon Sep 17 00:00:00 2001 From: Michael Ellerman Date: Thu, 16 Jun 2022 18:41:49 +1000 Subject: [PATCH 1/7] powerpc/mm: Move CMA reservations after initmem_init() After commit 11ac3e87ce09 ("mm: cma: use pageblock_order as the single alignment") there is an error at boot about the KVM CMA reservation failing, eg: kvm_cma_reserve: reserving 6553 MiB for global area cma: Failed to reserve 6553 MiB That makes it impossible to start KVM guests using the hash MMU with more than 2G of memory, because the VM is unable to allocate a large enough region for the hash page table, eg: $ qemu-system-ppc64 -enable-kvm -M pseries -m 4G ... qemu-system-ppc64: Failed to allocate KVM HPT of order 25: Cannot allocate memory Aneesh pointed out that this happens because when kvm_cma_reserve() is called, pageblock_order has not been initialised yet, and is still zero, causing the checks in cma_init_reserved_mem() against CMA_MIN_ALIGNMENT_PAGES to fail. Fix it by moving the call to kvm_cma_reserve() after initmem_init(). The pageblock_order is initialised in sparse_init() which is called from initmem_init(). Also move the hugetlb CMA reservation. Fixes: 11ac3e87ce09 ("mm: cma: use pageblock_order as the single alignment") Reviewed-by: Aneesh Kumar K.V Reviewed-by: Zi Yan Signed-off-by: Michael Ellerman Link: https://lore.kernel.org/r/20220616120033.1976732-1-mpe@ellerman.id.au --- arch/powerpc/kernel/setup-common.c | 13 +++++++------ 1 file changed, 7 insertions(+), 6 deletions(-) diff --git a/arch/powerpc/kernel/setup-common.c b/arch/powerpc/kernel/setup-common.c index eb0077b302e2..1a02629ec70b 100644 --- a/arch/powerpc/kernel/setup-common.c +++ b/arch/powerpc/kernel/setup-common.c @@ -935,12 +935,6 @@ void __init setup_arch(char **cmdline_p) /* Print various info about the machine that has been gathered so far. */ print_system_info(); - /* Reserve large chunks of memory for use by CMA for KVM. */ - kvm_cma_reserve(); - - /* Reserve large chunks of memory for us by CMA for hugetlb */ - gigantic_hugetlb_cma_reserve(); - klp_init_thread_info(&init_task); setup_initial_init_mm(_stext, _etext, _edata, _end); @@ -955,6 +949,13 @@ void __init setup_arch(char **cmdline_p) initmem_init(); + /* + * Reserve large chunks of memory for use by CMA for KVM and hugetlb. These must + * be called after initmem_init(), so that pageblock_order is initialised. + */ + kvm_cma_reserve(); + gigantic_hugetlb_cma_reserve(); + early_memtest(min_low_pfn << PAGE_SHIFT, max_low_pfn << PAGE_SHIFT); if (ppc_md.setup_arch) From 20a9689b3607456d92c6fb764501f6a95950b098 Mon Sep 17 00:00:00 2001 From: "Jason A. Donenfeld" Date: Sat, 11 Jun 2022 17:10:13 +0200 Subject: [PATCH 2/7] powerpc/microwatt: wire up rng during setup_arch() The platform's RNG must be available before random_init() in order to be useful for initial seeding, which in turn means that it needs to be called from setup_arch(), rather than from an init call. Fortunately, each platform already has a setup_arch function pointer, which means it's easy to wire this up. This commit also removes some noisy log messages that don't add much. Fixes: c25769fddaec ("powerpc/microwatt: Add support for hardware random number generator") Cc: stable@vger.kernel.org # v5.14+ Signed-off-by: Jason A. Donenfeld Reviewed-by: Christophe Leroy Signed-off-by: Michael Ellerman Link: https://lore.kernel.org/r/20220611151015.548325-2-Jason@zx2c4.com --- arch/powerpc/platforms/microwatt/microwatt.h | 7 +++++++ arch/powerpc/platforms/microwatt/rng.c | 10 +++------- arch/powerpc/platforms/microwatt/setup.c | 8 ++++++++ 3 files changed, 18 insertions(+), 7 deletions(-) create mode 100644 arch/powerpc/platforms/microwatt/microwatt.h diff --git a/arch/powerpc/platforms/microwatt/microwatt.h b/arch/powerpc/platforms/microwatt/microwatt.h new file mode 100644 index 000000000000..335417e95e66 --- /dev/null +++ b/arch/powerpc/platforms/microwatt/microwatt.h @@ -0,0 +1,7 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef _MICROWATT_H +#define _MICROWATT_H + +void microwatt_rng_init(void); + +#endif /* _MICROWATT_H */ diff --git a/arch/powerpc/platforms/microwatt/rng.c b/arch/powerpc/platforms/microwatt/rng.c index 7bc4d1cbfaf0..8ece87d005c8 100644 --- a/arch/powerpc/platforms/microwatt/rng.c +++ b/arch/powerpc/platforms/microwatt/rng.c @@ -11,6 +11,7 @@ #include #include #include +#include "microwatt.h" #define DARN_ERR 0xFFFFFFFFFFFFFFFFul @@ -29,7 +30,7 @@ static int microwatt_get_random_darn(unsigned long *v) return 1; } -static __init int rng_init(void) +void __init microwatt_rng_init(void) { unsigned long val; int i; @@ -37,12 +38,7 @@ static __init int rng_init(void) for (i = 0; i < 10; i++) { if (microwatt_get_random_darn(&val)) { ppc_md.get_random_seed = microwatt_get_random_darn; - return 0; + return; } } - - pr_warn("Unable to use DARN for get_random_seed()\n"); - - return -EIO; } -machine_subsys_initcall(, rng_init); diff --git a/arch/powerpc/platforms/microwatt/setup.c b/arch/powerpc/platforms/microwatt/setup.c index 0b02603bdb74..6b32539395a4 100644 --- a/arch/powerpc/platforms/microwatt/setup.c +++ b/arch/powerpc/platforms/microwatt/setup.c @@ -16,6 +16,8 @@ #include #include +#include "microwatt.h" + static void __init microwatt_init_IRQ(void) { xics_init(); @@ -32,10 +34,16 @@ static int __init microwatt_populate(void) } machine_arch_initcall(microwatt, microwatt_populate); +static void __init microwatt_setup_arch(void) +{ + microwatt_rng_init(); +} + define_machine(microwatt) { .name = "microwatt", .probe = microwatt_probe, .init_IRQ = microwatt_init_IRQ, + .setup_arch = microwatt_setup_arch, .progress = udbg_progress, .calibrate_decr = generic_calibrate_decr, }; From e561e472a3d441753bd012333b057f48fef1045b Mon Sep 17 00:00:00 2001 From: "Jason A. Donenfeld" Date: Sat, 11 Jun 2022 17:10:15 +0200 Subject: [PATCH 3/7] powerpc/pseries: wire up rng during setup_arch() The platform's RNG must be available before random_init() in order to be useful for initial seeding, which in turn means that it needs to be called from setup_arch(), rather than from an init call. Fortunately, each platform already has a setup_arch function pointer, which means it's easy to wire this up. This commit also removes some noisy log messages that don't add much. Fixes: a489043f4626 ("powerpc/pseries: Implement arch_get_random_long() based on H_RANDOM") Cc: stable@vger.kernel.org # v3.13+ Signed-off-by: Jason A. Donenfeld Reviewed-by: Christophe Leroy Signed-off-by: Michael Ellerman Link: https://lore.kernel.org/r/20220611151015.548325-4-Jason@zx2c4.com --- arch/powerpc/platforms/pseries/pseries.h | 2 ++ arch/powerpc/platforms/pseries/rng.c | 11 +++-------- arch/powerpc/platforms/pseries/setup.c | 1 + 3 files changed, 6 insertions(+), 8 deletions(-) diff --git a/arch/powerpc/platforms/pseries/pseries.h b/arch/powerpc/platforms/pseries/pseries.h index f5c916c839c9..1d75b7742ef0 100644 --- a/arch/powerpc/platforms/pseries/pseries.h +++ b/arch/powerpc/platforms/pseries/pseries.h @@ -122,4 +122,6 @@ void pseries_lpar_read_hblkrm_characteristics(void); static inline void pseries_lpar_read_hblkrm_characteristics(void) { } #endif +void pseries_rng_init(void); + #endif /* _PSERIES_PSERIES_H */ diff --git a/arch/powerpc/platforms/pseries/rng.c b/arch/powerpc/platforms/pseries/rng.c index 6268545947b8..6ddfdeaace9e 100644 --- a/arch/powerpc/platforms/pseries/rng.c +++ b/arch/powerpc/platforms/pseries/rng.c @@ -10,6 +10,7 @@ #include #include #include +#include "pseries.h" static int pseries_get_random_long(unsigned long *v) @@ -24,19 +25,13 @@ static int pseries_get_random_long(unsigned long *v) return 0; } -static __init int rng_init(void) +void __init pseries_rng_init(void) { struct device_node *dn; dn = of_find_compatible_node(NULL, NULL, "ibm,random"); if (!dn) - return -ENODEV; - - pr_info("Registering arch random hook.\n"); - + return; ppc_md.get_random_seed = pseries_get_random_long; - of_node_put(dn); - return 0; } -machine_subsys_initcall(pseries, rng_init); diff --git a/arch/powerpc/platforms/pseries/setup.c b/arch/powerpc/platforms/pseries/setup.c index afb074269b42..ee4f1db49515 100644 --- a/arch/powerpc/platforms/pseries/setup.c +++ b/arch/powerpc/platforms/pseries/setup.c @@ -839,6 +839,7 @@ static void __init pSeries_setup_arch(void) } ppc_md.pcibios_root_bridge_prepare = pseries_root_bridge_prepare; + pseries_rng_init(); } static void pseries_panic(char *str) From ec6d0dde71d760aa60316f8d1c9a1b0d99213529 Mon Sep 17 00:00:00 2001 From: "Naveen N. Rao" Date: Thu, 9 Jun 2022 16:03:28 +0530 Subject: [PATCH 4/7] powerpc: Enable execve syscall exit tracepoint On execve[at], we are zero'ing out most of the thread register state including gpr[0], which contains the syscall number. Due to this, we fail to trigger the syscall exit tracepoint properly. Fix this by retaining gpr[0] in the thread register state. Before this patch: # tail /sys/kernel/debug/tracing/trace cat-123 [000] ..... 61.449351: sys_execve(filename: 7fffa6b23448, argv: 7fffa6b233e0, envp: 7fffa6b233f8) cat-124 [000] ..... 62.428481: sys_execve(filename: 7fffa6b23448, argv: 7fffa6b233e0, envp: 7fffa6b233f8) echo-125 [000] ..... 65.813702: sys_execve(filename: 7fffa6b23378, argv: 7fffa6b233a0, envp: 7fffa6b233b0) echo-125 [000] ..... 65.822214: sys_execveat(fd: 0, filename: 1009ac48, argv: 7ffff65d0c98, envp: 7ffff65d0ca8, flags: 0) After this patch: # tail /sys/kernel/debug/tracing/trace cat-127 [000] ..... 100.416262: sys_execve(filename: 7fffa41b3448, argv: 7fffa41b33e0, envp: 7fffa41b33f8) cat-127 [000] ..... 100.418203: sys_execve -> 0x0 echo-128 [000] ..... 103.873968: sys_execve(filename: 7fffa41b3378, argv: 7fffa41b33a0, envp: 7fffa41b33b0) echo-128 [000] ..... 103.875102: sys_execve -> 0x0 echo-128 [000] ..... 103.882097: sys_execveat(fd: 0, filename: 1009ac48, argv: 7fffd10d2148, envp: 7fffd10d2158, flags: 0) echo-128 [000] ..... 103.883225: sys_execveat -> 0x0 Cc: stable@vger.kernel.org Signed-off-by: Naveen N. Rao Tested-by: Sumit Dubey2 Signed-off-by: Michael Ellerman Link: https://lore.kernel.org/r/20220609103328.41306-1-naveen.n.rao@linux.vnet.ibm.com --- arch/powerpc/kernel/process.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/powerpc/kernel/process.c b/arch/powerpc/kernel/process.c index ee0433809621..0fbda89cd1bb 100644 --- a/arch/powerpc/kernel/process.c +++ b/arch/powerpc/kernel/process.c @@ -1855,7 +1855,7 @@ void start_thread(struct pt_regs *regs, unsigned long start, unsigned long sp) tm_reclaim_current(0); #endif - memset(regs->gpr, 0, sizeof(regs->gpr)); + memset(®s->gpr[1], 0, sizeof(regs->gpr) - sizeof(regs->gpr[0])); regs->ctr = 0; regs->link = 0; regs->xer = 0; From 7bc08056a6dabc3a1442216daf527edf61ac24b6 Mon Sep 17 00:00:00 2001 From: Andrew Donnellan Date: Tue, 14 Jun 2022 23:49:52 +1000 Subject: [PATCH 5/7] powerpc/rtas: Allow ibm,platform-dump RTAS call with null buffer address Add a special case to block_rtas_call() to allow the ibm,platform-dump RTAS call through the RTAS filter if the buffer address is 0. According to PAPR, ibm,platform-dump is called with a null buffer address to notify the platform firmware that processing of a particular dump is finished. Without this, on a pseries machine with CONFIG_PPC_RTAS_FILTER enabled, an application such as rtas_errd that is attempting to retrieve a dump will encounter an error at the end of the retrieval process. Fixes: bd59380c5ba4 ("powerpc/rtas: Restrict RTAS requests from userspace") Cc: stable@vger.kernel.org Reported-by: Sathvika Vasireddy Signed-off-by: Andrew Donnellan Reviewed-by: Tyrel Datwyler Reviewed-by: Nathan Lynch Signed-off-by: Michael Ellerman Link: https://lore.kernel.org/r/20220614134952.156010-1-ajd@linux.ibm.com --- arch/powerpc/kernel/rtas.c | 11 ++++++++++- 1 file changed, 10 insertions(+), 1 deletion(-) diff --git a/arch/powerpc/kernel/rtas.c b/arch/powerpc/kernel/rtas.c index a6fce3106e02..693133972294 100644 --- a/arch/powerpc/kernel/rtas.c +++ b/arch/powerpc/kernel/rtas.c @@ -1071,7 +1071,7 @@ static struct rtas_filter rtas_filters[] __ro_after_init = { { "get-time-of-day", -1, -1, -1, -1, -1 }, { "ibm,get-vpd", -1, 0, -1, 1, 2 }, { "ibm,lpar-perftools", -1, 2, 3, -1, -1 }, - { "ibm,platform-dump", -1, 4, 5, -1, -1 }, + { "ibm,platform-dump", -1, 4, 5, -1, -1 }, /* Special cased */ { "ibm,read-slot-reset-state", -1, -1, -1, -1, -1 }, { "ibm,scan-log-dump", -1, 0, 1, -1, -1 }, { "ibm,set-dynamic-indicator", -1, 2, -1, -1, -1 }, @@ -1120,6 +1120,15 @@ static bool block_rtas_call(int token, int nargs, size = 1; end = base + size - 1; + + /* + * Special case for ibm,platform-dump - NULL buffer + * address is used to indicate end of dump processing + */ + if (!strcmp(f->name, "ibm,platform-dump") && + base == 0) + return false; + if (!in_rmo_buf(base, end)) goto err; } From ca5dabcff1df6bc8c413922b5fa63cc602858803 Mon Sep 17 00:00:00 2001 From: Christophe Leroy Date: Fri, 10 Jun 2022 17:43:43 +0200 Subject: [PATCH 6/7] powerpc/prom_init: Fix build failure with GCC_PLUGIN_STRUCTLEAK_BYREF_ALL and KASAN When CONFIG_KASAN is selected, we expect prom_init to use __memset() because it is too early to use memset(). But with CONFIG_GCC_PLUGIN_STRUCTLEAK_BYREF_ALL, the compiler adds calls to memset() to clear objects on stack, hence the following failure: PROMCHK arch/powerpc/kernel/prom_init_check Error: External symbol 'memset' referenced from prom_init.c make[2]: *** [arch/powerpc/kernel/Makefile:204 : arch/powerpc/kernel/prom_init_check] Erreur 1 prom_find_machine_type() is called from prom_init() and is called only once, so lets put compat[] in BSS instead of stack to avoid that. Signed-off-by: Christophe Leroy Signed-off-by: Michael Ellerman Link: https://lore.kernel.org/r/3802811f7cf94f730be44688539c01bba3a3b5c0.1654875808.git.christophe.leroy@csgroup.eu --- arch/powerpc/kernel/prom_init.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/powerpc/kernel/prom_init.c b/arch/powerpc/kernel/prom_init.c index 04694ec423f6..13d6cb188835 100644 --- a/arch/powerpc/kernel/prom_init.c +++ b/arch/powerpc/kernel/prom_init.c @@ -2302,7 +2302,7 @@ static void __init prom_init_stdout(void) static int __init prom_find_machine_type(void) { - char compat[256]; + static char compat[256] __prombss; int len, i = 0; #ifdef CONFIG_PPC64 phandle rtas; From f3eac426657d985b97c92fa5f7ae1d43f04721f3 Mon Sep 17 00:00:00 2001 From: "Jason A. Donenfeld" Date: Tue, 21 Jun 2022 16:08:49 +0200 Subject: [PATCH 7/7] powerpc/powernv: wire up rng during setup_arch The platform's RNG must be available before random_init() in order to be useful for initial seeding, which in turn means that it needs to be called from setup_arch(), rather than from an init call. Complicating things, however, is that POWER8 systems need some per-cpu state and kmalloc, which isn't available at this stage. So we split things up into an early phase and a later opportunistic phase. This commit also removes some noisy log messages that don't add much. Fixes: a4da0d50b2a0 ("powerpc: Implement arch_get_random_long/int() for powernv") Cc: stable@vger.kernel.org # v3.13+ Signed-off-by: Jason A. Donenfeld Reviewed-by: Christophe Leroy [mpe: Add of_node_put(), use pnv naming, minor change log editing] Signed-off-by: Michael Ellerman Link: https://lore.kernel.org/r/20220621140849.127227-1-Jason@zx2c4.com --- arch/powerpc/platforms/powernv/powernv.h | 2 + arch/powerpc/platforms/powernv/rng.c | 52 ++++++++++++++++-------- arch/powerpc/platforms/powernv/setup.c | 2 + 3 files changed, 40 insertions(+), 16 deletions(-) diff --git a/arch/powerpc/platforms/powernv/powernv.h b/arch/powerpc/platforms/powernv/powernv.h index e297bf4abfcb..866efdc103fd 100644 --- a/arch/powerpc/platforms/powernv/powernv.h +++ b/arch/powerpc/platforms/powernv/powernv.h @@ -42,4 +42,6 @@ ssize_t memcons_copy(struct memcons *mc, char *to, loff_t pos, size_t count); u32 __init memcons_get_size(struct memcons *mc); struct memcons *__init memcons_init(struct device_node *node, const char *mc_prop_name); +void pnv_rng_init(void); + #endif /* _POWERNV_H */ diff --git a/arch/powerpc/platforms/powernv/rng.c b/arch/powerpc/platforms/powernv/rng.c index e3d44b36ae98..463c78c52cc5 100644 --- a/arch/powerpc/platforms/powernv/rng.c +++ b/arch/powerpc/platforms/powernv/rng.c @@ -17,6 +17,7 @@ #include #include #include +#include "powernv.h" #define DARN_ERR 0xFFFFFFFFFFFFFFFFul @@ -28,7 +29,6 @@ struct powernv_rng { static DEFINE_PER_CPU(struct powernv_rng *, powernv_rng); - int powernv_hwrng_present(void) { struct powernv_rng *rng; @@ -98,9 +98,6 @@ static int __init initialise_darn(void) return 0; } } - - pr_warn("Unable to use DARN for get_random_seed()\n"); - return -EIO; } @@ -163,32 +160,55 @@ static __init int rng_create(struct device_node *dn) rng_init_per_cpu(rng, dn); - pr_info_once("Registering arch random hook.\n"); - ppc_md.get_random_seed = powernv_get_random_long; return 0; } -static __init int rng_init(void) +static int __init pnv_get_random_long_early(unsigned long *v) { struct device_node *dn; - int rc; + + if (!slab_is_available()) + return 0; + + if (cmpxchg(&ppc_md.get_random_seed, pnv_get_random_long_early, + NULL) != pnv_get_random_long_early) + return 0; for_each_compatible_node(dn, NULL, "ibm,power-rng") { - rc = rng_create(dn); - if (rc) { - pr_err("Failed creating rng for %pOF (%d).\n", - dn, rc); + if (rng_create(dn)) continue; - } - /* Create devices for hwrng driver */ of_platform_device_create(dn, NULL, NULL); } - initialise_darn(); + if (!ppc_md.get_random_seed) + return 0; + return ppc_md.get_random_seed(v); +} +void __init pnv_rng_init(void) +{ + struct device_node *dn; + + /* Prefer darn over the rest. */ + if (!initialise_darn()) + return; + + dn = of_find_compatible_node(NULL, NULL, "ibm,power-rng"); + if (dn) + ppc_md.get_random_seed = pnv_get_random_long_early; + + of_node_put(dn); +} + +static int __init pnv_rng_late_init(void) +{ + unsigned long v; + /* In case it wasn't called during init for some other reason. */ + if (ppc_md.get_random_seed == pnv_get_random_long_early) + pnv_get_random_long_early(&v); return 0; } -machine_subsys_initcall(powernv, rng_init); +machine_subsys_initcall(powernv, pnv_rng_late_init); diff --git a/arch/powerpc/platforms/powernv/setup.c b/arch/powerpc/platforms/powernv/setup.c index 824c3ad7a0fa..dac545aa0308 100644 --- a/arch/powerpc/platforms/powernv/setup.c +++ b/arch/powerpc/platforms/powernv/setup.c @@ -203,6 +203,8 @@ static void __init pnv_setup_arch(void) pnv_check_guarded_cores(); /* XXX PMCS */ + + pnv_rng_init(); } static void __init pnv_init(void)