From af22bbe1f4a514c80b89a27252beef033168f4e9 Mon Sep 17 00:00:00 2001 From: Jiri Pirko Date: Tue, 16 Jul 2024 13:35:48 +0200 Subject: virtio: create admin queues alongside other virtqueues Admin virtqueue is just another virtqueue nothing that special about it. The current implementation treats it somehow separate though in terms of creation and deletion. Unify the admin virtqueue creation and deletion flows to be aligned with the rest of virtqueues, creating it from vp_find_vqs_*() helpers. Let the admin virtqueue to be deleted by vp_del_vqs() as the rest. Call vp_find_one_vq_msix() with slow_path argument being "true" to make sure that in case of limited interrupt vectors the config vector is used for admin queue. Signed-off-by: Jiri Pirko Message-Id: <20240716113552.80599-10-jiri@resnulli.us> Signed-off-by: Michael S. Tsirkin --- include/linux/virtio_config.h | 4 ---- 1 file changed, 4 deletions(-) (limited to 'include/linux') diff --git a/include/linux/virtio_config.h b/include/linux/virtio_config.h index ab4b9a3fef6b..169c7d367fac 100644 --- a/include/linux/virtio_config.h +++ b/include/linux/virtio_config.h @@ -104,8 +104,6 @@ struct virtqueue_info { * Returns 0 on success or error status * If disable_vq_and_reset is set, then enable_vq_after_reset must also be * set. - * @create_avq: create admin virtqueue resource. - * @destroy_avq: destroy admin virtqueue resource. */ struct virtio_config_ops { void (*get)(struct virtio_device *vdev, unsigned offset, @@ -133,8 +131,6 @@ struct virtio_config_ops { struct virtio_shm_region *region, u8 id); int (*disable_vq_and_reset)(struct virtqueue *vq); int (*enable_vq_after_reset)(struct virtqueue *vq); - int (*create_avq)(struct virtio_device *vdev); - void (*destroy_avq)(struct virtio_device *vdev); }; /* If driver didn't advertise the feature, it will never appear. */ -- cgit v1.2.3 From 4c3b54af907e709609d3d8beca92d65e2f0cfd83 Mon Sep 17 00:00:00 2001 From: Jiri Pirko Date: Tue, 16 Jul 2024 13:35:51 +0200 Subject: virtio_pci_modern: use completion instead of busy loop to wait on admin cmd result Currently, the code waits in a busy loop on every admin virtqueue issued command to get a reply. That prevents callers from issuing multiple commands in parallel. To overcome this limitation, introduce a virtqueue event callback for admin virtqueue. For every issued command, use completion mechanism to wait on a reply. In the event callback, trigger the completion is done for every incoming reply. Alongside with that, introduce a spin lock to protect the admin virtqueue operations. Signed-off-by: Jiri Pirko Message-Id: <20240716113552.80599-13-jiri@resnulli.us> Signed-off-by: Michael S. Tsirkin --- include/linux/virtio.h | 3 +++ 1 file changed, 3 insertions(+) (limited to 'include/linux') diff --git a/include/linux/virtio.h b/include/linux/virtio.h index 96fea920873b..999ff5934392 100644 --- a/include/linux/virtio.h +++ b/include/linux/virtio.h @@ -10,6 +10,7 @@ #include #include #include +#include /** * struct virtqueue - a queue to register buffers for sending or receiving. @@ -109,6 +110,8 @@ struct virtio_admin_cmd { __le64 group_member_id; struct scatterlist *data_sg; struct scatterlist *result_sg; + struct completion completion; + int ret; }; /** -- cgit v1.2.3 From 564429a6bd8d26065b2cccffcaa9485359f74de7 Mon Sep 17 00:00:00 2001 From: Paolo Bonzini Date: Thu, 11 Jul 2024 18:27:47 -0400 Subject: KVM: rename CONFIG_HAVE_KVM_GMEM_* to CONFIG_HAVE_KVM_ARCH_GMEM_* Add "ARCH" to the symbols; shortly, the "prepare" phase will include both the arch-independent step to clear out contents left in the page by the host, and the arch-dependent step enabled by CONFIG_HAVE_KVM_GMEM_PREPARE. For consistency do the same for CONFIG_HAVE_KVM_GMEM_INVALIDATE as well. Reviewed-by: Michael Roth Signed-off-by: Paolo Bonzini --- include/linux/kvm_host.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'include/linux') diff --git a/include/linux/kvm_host.h b/include/linux/kvm_host.h index 689e8be873a7..344d90771844 100644 --- a/include/linux/kvm_host.h +++ b/include/linux/kvm_host.h @@ -2445,7 +2445,7 @@ static inline int kvm_gmem_get_pfn(struct kvm *kvm, } #endif /* CONFIG_KVM_PRIVATE_MEM */ -#ifdef CONFIG_HAVE_KVM_GMEM_PREPARE +#ifdef CONFIG_HAVE_KVM_ARCH_GMEM_PREPARE int kvm_arch_gmem_prepare(struct kvm *kvm, gfn_t gfn, kvm_pfn_t pfn, int max_order); bool kvm_arch_gmem_prepare_needed(struct kvm *kvm); #endif @@ -2477,7 +2477,7 @@ typedef int (*kvm_gmem_populate_cb)(struct kvm *kvm, gfn_t gfn, kvm_pfn_t pfn, long kvm_gmem_populate(struct kvm *kvm, gfn_t gfn, void __user *src, long npages, kvm_gmem_populate_cb post_populate, void *opaque); -#ifdef CONFIG_HAVE_KVM_GMEM_INVALIDATE +#ifdef CONFIG_HAVE_KVM_ARCH_GMEM_INVALIDATE void kvm_arch_gmem_invalidate(kvm_pfn_t start, kvm_pfn_t end); #endif -- cgit v1.2.3 From 7239ed74677af143857d1a96d402476446a0995a Mon Sep 17 00:00:00 2001 From: Paolo Bonzini Date: Thu, 11 Jul 2024 18:27:51 -0400 Subject: KVM: remove kvm_arch_gmem_prepare_needed() It is enough to return 0 if a guest need not do any preparation. This is in fact how sev_gmem_prepare() works for non-SNP guests, and it extends naturally to Intel hosts: the x86 callback for gmem_prepare is optional and returns 0 if not defined. Reviewed-by: Michael Roth Signed-off-by: Paolo Bonzini --- include/linux/kvm_host.h | 1 - 1 file changed, 1 deletion(-) (limited to 'include/linux') diff --git a/include/linux/kvm_host.h b/include/linux/kvm_host.h index 344d90771844..45373d42f314 100644 --- a/include/linux/kvm_host.h +++ b/include/linux/kvm_host.h @@ -2447,7 +2447,6 @@ static inline int kvm_gmem_get_pfn(struct kvm *kvm, #ifdef CONFIG_HAVE_KVM_ARCH_GMEM_PREPARE int kvm_arch_gmem_prepare(struct kvm *kvm, gfn_t gfn, kvm_pfn_t pfn, int max_order); -bool kvm_arch_gmem_prepare_needed(struct kvm *kvm); #endif /** -- cgit v1.2.3 From 4b5f67120a88c713b82907d55a767693382e9e9d Mon Sep 17 00:00:00 2001 From: Paolo Bonzini Date: Thu, 11 Jul 2024 18:27:54 -0400 Subject: KVM: extend kvm_range_has_memory_attributes() to check subset of attributes While currently there is no other attribute than KVM_MEMORY_ATTRIBUTE_PRIVATE, KVM code such as kvm_mem_is_private() is written to expect their existence. Allow using kvm_range_has_memory_attributes() as a multi-page version of kvm_mem_is_private(), without it breaking later when more attributes are introduced. Reviewed-by: Michael Roth Signed-off-by: Paolo Bonzini --- include/linux/kvm_host.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'include/linux') diff --git a/include/linux/kvm_host.h b/include/linux/kvm_host.h index 45373d42f314..c223b97df03e 100644 --- a/include/linux/kvm_host.h +++ b/include/linux/kvm_host.h @@ -2414,7 +2414,7 @@ static inline unsigned long kvm_get_memory_attributes(struct kvm *kvm, gfn_t gfn } bool kvm_range_has_memory_attributes(struct kvm *kvm, gfn_t start, gfn_t end, - unsigned long attrs); + unsigned long mask, unsigned long attrs); bool kvm_arch_pre_set_memory_attributes(struct kvm *kvm, struct kvm_gfn_range *range); bool kvm_arch_post_set_memory_attributes(struct kvm *kvm, -- cgit v1.2.3 From e4ee5447927377c55777b73fe497a2455a25f948 Mon Sep 17 00:00:00 2001 From: Paolo Bonzini Date: Thu, 11 Jul 2024 18:27:55 -0400 Subject: KVM: guest_memfd: let kvm_gmem_populate() operate only on private gfns This check is currently performed by sev_gmem_post_populate(), but it applies to all callers of kvm_gmem_populate(): the point of the function is that the memory is being encrypted and some work has to be done on all the gfns in order to encrypt them. Therefore, check the KVM_MEMORY_ATTRIBUTE_PRIVATE attribute prior to invoking the callback, and stop the operation if a shared page is encountered. Because CONFIG_KVM_PRIVATE_MEM in principle does not require attributes, this makes kvm_gmem_populate() depend on CONFIG_KVM_GENERIC_PRIVATE_MEM (which does require them). Reviewed-by: Michael Roth Signed-off-by: Paolo Bonzini --- include/linux/kvm_host.h | 2 ++ 1 file changed, 2 insertions(+) (limited to 'include/linux') diff --git a/include/linux/kvm_host.h b/include/linux/kvm_host.h index c223b97df03e..79a6b1a63027 100644 --- a/include/linux/kvm_host.h +++ b/include/linux/kvm_host.h @@ -2449,6 +2449,7 @@ static inline int kvm_gmem_get_pfn(struct kvm *kvm, int kvm_arch_gmem_prepare(struct kvm *kvm, gfn_t gfn, kvm_pfn_t pfn, int max_order); #endif +#ifdef CONFIG_KVM_GENERIC_PRIVATE_MEM /** * kvm_gmem_populate() - Populate/prepare a GPA range with guest data * @@ -2475,6 +2476,7 @@ typedef int (*kvm_gmem_populate_cb)(struct kvm *kvm, gfn_t gfn, kvm_pfn_t pfn, long kvm_gmem_populate(struct kvm *kvm, gfn_t gfn, void __user *src, long npages, kvm_gmem_populate_cb post_populate, void *opaque); +#endif #ifdef CONFIG_HAVE_KVM_ARCH_GMEM_INVALIDATE void kvm_arch_gmem_invalidate(kvm_pfn_t start, kvm_pfn_t end); -- cgit v1.2.3 From 1a251f52cfdc417c84411a056bc142cbd77baef4 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Sun, 28 Jul 2024 15:49:18 -0700 Subject: minmax: make generic MIN() and MAX() macros available everywhere This just standardizes the use of MIN() and MAX() macros, with the very traditional semantics. The goal is to use these for C constant expressions and for top-level / static initializers, and so be able to simplify the min()/max() macros. These macro names were used by various kernel code - they are very traditional, after all - and all such users have been fixed up, with a few different approaches: - trivial duplicated macro definitions have been removed Note that 'trivial' here means that it's obviously kernel code that already included all the major kernel headers, and thus gets the new generic MIN/MAX macros automatically. - non-trivial duplicated macro definitions are guarded with #ifndef This is the "yes, they define their own versions, but no, the include situation is not entirely obvious, and maybe they don't get the generic version automatically" case. - strange use case #1 A couple of drivers decided that the way they want to describe their versioning is with #define MAJ 1 #define MIN 2 #define DRV_VERSION __stringify(MAJ) "." __stringify(MIN) which adds zero value and I just did my Alexander the Great impersonation, and rewrote that pointless Gordian knot as #define DRV_VERSION "1.2" instead. - strange use case #2 A couple of drivers thought that it's a good idea to have a random 'MIN' or 'MAX' define for a value or index into a table, rather than the traditional macro that takes arguments. These values were re-written as C enum's instead. The new function-line macros only expand when followed by an open parenthesis, and thus don't clash with enum use. Happily, there weren't really all that many of these cases, and a lot of users already had the pattern of using '#ifndef' guarding (or in one case just using '#undef MIN') before defining their own private version that does the same thing. I left such cases alone. Cc: David Laight Cc: Lorenzo Stoakes Signed-off-by: Linus Torvalds --- include/linux/minmax.h | 2 ++ 1 file changed, 2 insertions(+) (limited to 'include/linux') diff --git a/include/linux/minmax.h b/include/linux/minmax.h index 9c2848abc804..fc384714da45 100644 --- a/include/linux/minmax.h +++ b/include/linux/minmax.h @@ -277,6 +277,8 @@ static inline bool in_range32(u32 val, u32 start, u32 len) * Use these carefully: no type checking, and uses the arguments * multiple times. Use for obvious constants only. */ +#define MIN(a,b) __cmp(min,a,b) +#define MAX(a,b) __cmp(max,a,b) #define MIN_T(type,a,b) __cmp(min,(type)(a),(type)(b)) #define MAX_T(type,a,b) __cmp(max,(type)(a),(type)(b)) -- cgit v1.2.3 From dc1c8034e31b14a2e5e212104ec508aec44ce1b9 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Sun, 28 Jul 2024 20:24:12 -0700 Subject: minmax: simplify min()/max()/clamp() implementation Now that we no longer have any C constant expression contexts (ie array size declarations or static initializers) that use min() or max(), we can simpify the implementation by not having to worry about the result staying as a C constant expression. So now we can unconditionally just use temporary variables of the right type, and get rid of the excessive expansion that used to come from the use of __builtin_choose_expr(__is_constexpr(...), .. to pick the specialized code for constant expressions. Another expansion simplification is to pass the temporary variables (in addition to the original expression) to our __types_ok() macro. That may superficially look like it complicates the macro, but when we only want the type of the expression, expanding the temporary variable names is much simpler and smaller than expanding the potentially complicated original expression. As a result, on my machine, doing a $ time make drivers/staging/media/atomisp/pci/isp/kernels/ynr/ynr_1.0/ia_css_ynr.host.i goes from real 0m16.621s user 0m15.360s sys 0m1.221s to real 0m2.532s user 0m2.091s sys 0m0.452s because the token expansion goes down dramatically. In particular, the longest line expansion (which was line 71 of that 'ia_css_ynr.host.c' file) shrinks from 23,338kB (yes, 23MB for one single line) to "just" 1,444kB (now "only" 1.4MB). And yes, that line is still the line from hell, because it's doing multiple levels of "min()/max()" expansion thanks to some of them being hidden inside the uDIGIT_FITTING() macro. Lorenzo has a nice cleanup patch that makes that driver use inline functions instead of macros for sDIGIT_FITTING() and uDIGIT_FITTING(), which will fix that line once and for all, but the 16-fold reduction in this case does show why we need to simplify these helpers. Cc: David Laight Cc: Lorenzo Stoakes Signed-off-by: Linus Torvalds --- include/linux/minmax.h | 43 ++++++++++++++++++++----------------------- 1 file changed, 20 insertions(+), 23 deletions(-) (limited to 'include/linux') diff --git a/include/linux/minmax.h b/include/linux/minmax.h index fc384714da45..e3e4353df983 100644 --- a/include/linux/minmax.h +++ b/include/linux/minmax.h @@ -35,10 +35,10 @@ #define __is_noneg_int(x) \ (__builtin_choose_expr(__is_constexpr(x) && __is_signed(x), x, -1) >= 0) -#define __types_ok(x, y) \ - (__is_signed(x) == __is_signed(y) || \ - __is_signed((x) + 0) == __is_signed((y) + 0) || \ - __is_noneg_int(x) || __is_noneg_int(y)) +#define __types_ok(x, y, ux, uy) \ + (__is_signed(ux) == __is_signed(uy) || \ + __is_signed((ux) + 0) == __is_signed((uy) + 0) || \ + __is_noneg_int(x) || __is_noneg_int(y)) #define __cmp_op_min < #define __cmp_op_max > @@ -51,34 +51,31 @@ #define __cmp_once(op, type, x, y) \ __cmp_once_unique(op, type, x, y, __UNIQUE_ID(x_), __UNIQUE_ID(y_)) -#define __careful_cmp_once(op, x, y) ({ \ - static_assert(__types_ok(x, y), \ +#define __careful_cmp_once(op, x, y, ux, uy) ({ \ + __auto_type ux = (x); __auto_type uy = (y); \ + static_assert(__types_ok(x, y, ux, uy), \ #op "(" #x ", " #y ") signedness error, fix types or consider u" #op "() before " #op "_t()"); \ - __cmp_once(op, __auto_type, x, y); }) + __cmp(op, ux, uy); }) -#define __careful_cmp(op, x, y) \ - __builtin_choose_expr(__is_constexpr((x) - (y)), \ - __cmp(op, x, y), __careful_cmp_once(op, x, y)) +#define __careful_cmp(op, x, y) \ + __careful_cmp_once(op, x, y, __UNIQUE_ID(x_), __UNIQUE_ID(y_)) #define __clamp(val, lo, hi) \ ((val) >= (hi) ? (hi) : ((val) <= (lo) ? (lo) : (val))) -#define __clamp_once(val, lo, hi, unique_val, unique_lo, unique_hi) ({ \ - typeof(val) unique_val = (val); \ - typeof(lo) unique_lo = (lo); \ - typeof(hi) unique_hi = (hi); \ +#define __clamp_once(val, lo, hi, uval, ulo, uhi) ({ \ + __auto_type uval = (val); \ + __auto_type ulo = (lo); \ + __auto_type uhi = (hi); \ static_assert(__builtin_choose_expr(__is_constexpr((lo) > (hi)), \ (lo) <= (hi), true), \ "clamp() low limit " #lo " greater than high limit " #hi); \ - static_assert(__types_ok(val, lo), "clamp() 'lo' signedness error"); \ - static_assert(__types_ok(val, hi), "clamp() 'hi' signedness error"); \ - __clamp(unique_val, unique_lo, unique_hi); }) - -#define __careful_clamp(val, lo, hi) ({ \ - __builtin_choose_expr(__is_constexpr((val) - (lo) + (hi)), \ - __clamp(val, lo, hi), \ - __clamp_once(val, lo, hi, __UNIQUE_ID(__val), \ - __UNIQUE_ID(__lo), __UNIQUE_ID(__hi))); }) + static_assert(__types_ok(uval, lo, uval, ulo), "clamp() 'lo' signedness error"); \ + static_assert(__types_ok(uval, hi, uval, uhi), "clamp() 'hi' signedness error"); \ + __clamp(uval, ulo, uhi); }) + +#define __careful_clamp(val, lo, hi) \ + __clamp_once(val, lo, hi, __UNIQUE_ID(v_), __UNIQUE_ID(l_), __UNIQUE_ID(h_)) /** * min - return minimum of two values of the same or compatible types -- cgit v1.2.3 From 2accfdb7eff65f390c4308b0e9cb7c3fe48ad63c Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Mon, 29 Jul 2024 10:58:28 -0700 Subject: profiling: attempt to remove per-cpu profile flip buffer This is the really old legacy kernel profiling code, which has long since been obviated by "real profiling" (ie 'prof' and company), and mainly remains as a source of syzbot reports. There are anecdotal reports that people still use it for boot-time profiling, but it's unlikely that such use would care about the old NUMA optimizations in this code from 2004 (commit ad02973d42: "profile: 512x Altix timer interrupt livelock fix" in the BK import archive at [1]) So in order to head off future syzbot reports, let's try to simplify this code and get rid of the per-cpu profile buffers that are quite a large portion of the complexity footprint of this thing (including CPU hotplug callbacks etc). It's unlikely anybody will actually notice, or possibly, as Thomas put it: "Only people who indulge in nostalgia will notice :)". That said, if it turns out that this code is actually actively used by somebody, we can always revert this removal. Thus the "attempt" in the summary line. [ Note: in a small nod to "the profiling code can cause NUMA problems", this also removes the "increment the last entry in the profiling array on any unknown hits" logic. That would account any program counter in a module to that single counter location, and might exacerbate any NUMA cacheline bouncing issues ] Link: https://lore.kernel.org/all/CAHk-=wgs52BxT4Zjmjz8aNvHWKxf5_ThBY4bYL1Y6CTaNL2dTw@mail.gmail.com/ Link: https://git.kernel.org/pub/scm/linux/kernel/git/tglx/history.git [1] Cc: Thomas Gleixner Cc: Tetsuo Handa Signed-off-by: Linus Torvalds --- include/linux/cpuhotplug.h | 1 - 1 file changed, 1 deletion(-) (limited to 'include/linux') diff --git a/include/linux/cpuhotplug.h b/include/linux/cpuhotplug.h index 51ba681b915a..affdd890899e 100644 --- a/include/linux/cpuhotplug.h +++ b/include/linux/cpuhotplug.h @@ -100,7 +100,6 @@ enum cpuhp_state { CPUHP_WORKQUEUE_PREP, CPUHP_POWER_NUMA_PREPARE, CPUHP_HRTIMERS_PREPARE, - CPUHP_PROFILE_PREPARE, CPUHP_X2APIC_PREPARE, CPUHP_SMPCFD_PREPARE, CPUHP_RELAY_PREPARE, -- cgit v1.2.3 From 22f5468731491e53356ba7c028f0fdea20b18e2c Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Tue, 30 Jul 2024 10:36:47 -0700 Subject: minmax: improve macro expansion and type checking This clarifies the rules for min()/max()/clamp() type checking and makes them a much more efficient macro expansion. In particular, we now look at the type and range of the inputs to see whether they work together, generating a mask of acceptable comparisons, and then just verifying that the inputs have a shared case: - an expression with a signed type can be used for (1) signed comparisons (2) unsigned comparisons if it is statically known to have a non-negative value - an expression with an unsigned type can be used for (3) unsigned comparison (4) signed comparisons if the type is smaller than 'int' and thus the C integer promotion rules will make it signed anyway Here rule (1) and (3) are obvious, and rule (2) is important in order to allow obvious trivial constants to be used together with unsigned values. Rule (4) is not necessarily a good idea, but matches what we used to do, and we have extant cases of this situation in the kernel. Notably with bcachefs having an expression like min(bch2_bucket_sectors_dirty(a), ca->mi.bucket_size) where bch2_bucket_sectors_dirty() returns an 's64', and 'ca->mi.bucket_size' is of type 'u16'. Technically that bcachefs comparison is clearly sensible on a C type level, because the 'u16' will go through the normal C integer promotion, and become 'int', and then we're comparing two signed values and everything looks sane. However, it's not entirely clear that a 'min(s64,u16)' operation makes a lot of conceptual sense, and it's possible that we will remove rule (4). After all, the _reason_ we have these complicated type checks is exactly that the C type promotion rules are not very intuitive. But at least for now the rule is in place for backwards compatibility. Also note that rule (2) existed before, but is hugely relaxed by this commit. It used to be true only for the simplest compile-time non-negative integer constants. The new macro model will allow cases where the compiler can trivially see that an expression is non-negative even if it isn't necessarily a constant. For example, the amdgpu driver does min_t(size_t, sizeof(fru_info->serial), pia[addr] & 0x3F)); because our old 'min()' macro would see that 'pia[addr] & 0x3F' is of type 'int' and clearly not a C constant expression, so doing a 'min()' with a 'size_t' is a signedness violation. Our new 'min()' macro still sees that 'pia[addr] & 0x3F' is of type 'int', but is smart enough to also see that it is clearly non-negative, and thus would allow that case without any complaints. Cc: Arnd Bergmann Cc: David Laight Cc: Lorenzo Stoakes Signed-off-by: Linus Torvalds --- include/linux/compiler.h | 9 ++++++ include/linux/minmax.h | 74 ++++++++++++++++++++++++++++++++++++++---------- 2 files changed, 68 insertions(+), 15 deletions(-) (limited to 'include/linux') diff --git a/include/linux/compiler.h b/include/linux/compiler.h index 2594553bb30b..2df665fa2964 100644 --- a/include/linux/compiler.h +++ b/include/linux/compiler.h @@ -296,6 +296,15 @@ static inline void *offset_to_ptr(const int *off) #define is_signed_type(type) (((type)(-1)) < (__force type)1) #define is_unsigned_type(type) (!is_signed_type(type)) +/* + * Useful shorthand for "is this condition known at compile-time?" + * + * Note that the condition may involve non-constant values, + * but the compiler may know enough about the details of the + * values to determine that the condition is statically true. + */ +#define statically_true(x) (__builtin_constant_p(x) && (x)) + /* * This is needed in functions which generate the stack canary, see * arch/x86/kernel/smpboot.c::start_secondary() for an example. diff --git a/include/linux/minmax.h b/include/linux/minmax.h index e3e4353df983..41da6f85a407 100644 --- a/include/linux/minmax.h +++ b/include/linux/minmax.h @@ -26,19 +26,63 @@ #define __typecheck(x, y) \ (!!(sizeof((typeof(x) *)1 == (typeof(y) *)1))) -/* is_signed_type() isn't a constexpr for pointer types */ -#define __is_signed(x) \ - __builtin_choose_expr(__is_constexpr(is_signed_type(typeof(x))), \ - is_signed_type(typeof(x)), 0) +/* + * __sign_use for integer expressions: + * bit #0 set if ok for unsigned comparisons + * bit #1 set if ok for signed comparisons + * + * In particular, statically non-negative signed integer + * expressions are ok for both. + * + * NOTE! Unsigned types smaller than 'int' are implicitly + * converted to 'int' in expressions, and are accepted for + * signed conversions for now. This is debatable. + * + * Note that 'x' is the original expression, and 'ux' is + * the unique variable that contains the value. + * + * We use 'ux' for pure type checking, and 'x' for when + * we need to look at the value (but without evaluating + * it for side effects! Careful to only ever evaluate it + * with sizeof() or __builtin_constant_p() etc). + * + * Pointers end up being checked by the normal C type + * rules at the actual comparison, and these expressions + * only need to be careful to not cause warnings for + * pointer use. + */ +#define __signed_type_use(x,ux) (2+__is_nonneg(x,ux)) +#define __unsigned_type_use(x,ux) (1+2*(sizeof(ux)<4)) +#define __sign_use(x,ux) (is_signed_type(typeof(ux))? \ + __signed_type_use(x,ux):__unsigned_type_use(x,ux)) + +/* + * To avoid warnings about casting pointers to integers + * of different sizes, we need that special sign type. + * + * On 64-bit we can just always use 'long', since any + * integer or pointer type can just be cast to that. + * + * This does not work for 128-bit signed integers since + * the cast would truncate them, but we do not use s128 + * types in the kernel (we do use 'u128', but they will + * be handled by the !is_signed_type() case). + * + * NOTE! The cast is there only to avoid any warnings + * from when values that aren't signed integer types. + */ +#ifdef CONFIG_64BIT + #define __signed_type(ux) long +#else + #define __signed_type(ux) typeof(__builtin_choose_expr(sizeof(ux)>4,1LL,1L)) +#endif +#define __is_nonneg(x,ux) statically_true((__signed_type(ux))(x)>=0) -/* True for a non-negative signed int constant */ -#define __is_noneg_int(x) \ - (__builtin_choose_expr(__is_constexpr(x) && __is_signed(x), x, -1) >= 0) +#define __types_ok(x,y,ux,uy) \ + (__sign_use(x,ux) & __sign_use(y,uy)) -#define __types_ok(x, y, ux, uy) \ - (__is_signed(ux) == __is_signed(uy) || \ - __is_signed((ux) + 0) == __is_signed((uy) + 0) || \ - __is_noneg_int(x) || __is_noneg_int(y)) +#define __types_ok3(x,y,z,ux,uy,uz) \ + (__sign_use(x,ux) & __sign_use(y,uy) & __sign_use(z,uz)) #define __cmp_op_min < #define __cmp_op_max > @@ -53,8 +97,8 @@ #define __careful_cmp_once(op, x, y, ux, uy) ({ \ __auto_type ux = (x); __auto_type uy = (y); \ - static_assert(__types_ok(x, y, ux, uy), \ - #op "(" #x ", " #y ") signedness error, fix types or consider u" #op "() before " #op "_t()"); \ + BUILD_BUG_ON_MSG(!__types_ok(x,y,ux,uy), \ + #op"("#x", "#y") signedness error"); \ __cmp(op, ux, uy); }) #define __careful_cmp(op, x, y) \ @@ -70,8 +114,8 @@ static_assert(__builtin_choose_expr(__is_constexpr((lo) > (hi)), \ (lo) <= (hi), true), \ "clamp() low limit " #lo " greater than high limit " #hi); \ - static_assert(__types_ok(uval, lo, uval, ulo), "clamp() 'lo' signedness error"); \ - static_assert(__types_ok(uval, hi, uval, uhi), "clamp() 'hi' signedness error"); \ + BUILD_BUG_ON_MSG(!__types_ok3(val,lo,hi,uval,ulo,uhi), \ + "clamp("#val", "#lo", "#hi") signedness error"); \ __clamp(uval, ulo, uhi); }) #define __careful_clamp(val, lo, hi) \ -- cgit v1.2.3 From 89add40066f9ed9abe5f7f886fe5789ff7e0c50e Mon Sep 17 00:00:00 2001 From: Willem de Bruijn Date: Mon, 29 Jul 2024 16:10:12 -0400 Subject: net: drop bad gso csum_start and offset in virtio_net_hdr Tighten csum_start and csum_offset checks in virtio_net_hdr_to_skb for GSO packets. The function already checks that a checksum requested with VIRTIO_NET_HDR_F_NEEDS_CSUM is in skb linear. But for GSO packets this might not hold for segs after segmentation. Syzkaller demonstrated to reach this warning in skb_checksum_help offset = skb_checksum_start_offset(skb); ret = -EINVAL; if (WARN_ON_ONCE(offset >= skb_headlen(skb))) By injecting a TSO packet: WARNING: CPU: 1 PID: 3539 at net/core/dev.c:3284 skb_checksum_help+0x3d0/0x5b0 ip_do_fragment+0x209/0x1b20 net/ipv4/ip_output.c:774 ip_finish_output_gso net/ipv4/ip_output.c:279 [inline] __ip_finish_output+0x2bd/0x4b0 net/ipv4/ip_output.c:301 iptunnel_xmit+0x50c/0x930 net/ipv4/ip_tunnel_core.c:82 ip_tunnel_xmit+0x2296/0x2c70 net/ipv4/ip_tunnel.c:813 __gre_xmit net/ipv4/ip_gre.c:469 [inline] ipgre_xmit+0x759/0xa60 net/ipv4/ip_gre.c:661 __netdev_start_xmit include/linux/netdevice.h:4850 [inline] netdev_start_xmit include/linux/netdevice.h:4864 [inline] xmit_one net/core/dev.c:3595 [inline] dev_hard_start_xmit+0x261/0x8c0 net/core/dev.c:3611 __dev_queue_xmit+0x1b97/0x3c90 net/core/dev.c:4261 packet_snd net/packet/af_packet.c:3073 [inline] The geometry of the bad input packet at tcp_gso_segment: [ 52.003050][ T8403] skb len=12202 headroom=244 headlen=12093 tailroom=0 [ 52.003050][ T8403] mac=(168,24) mac_len=24 net=(192,52) trans=244 [ 52.003050][ T8403] shinfo(txflags=0 nr_frags=1 gso(size=1552 type=3 segs=0)) [ 52.003050][ T8403] csum(0x60000c7 start=199 offset=1536 ip_summed=3 complete_sw=0 valid=0 level=0) Mitigate with stricter input validation. csum_offset: for GSO packets, deduce the correct value from gso_type. This is already done for USO. Extend it to TSO. Let UFO be: udp[46]_ufo_fragment ignores these fields and always computes the checksum in software. csum_start: finding the real offset requires parsing to the transport header. Do not add a parser, use existing segmentation parsing. Thanks to SKB_GSO_DODGY, that also catches bad packets that are hw offloaded. Again test both TSO and USO. Do not test UFO for the above reason, and do not test UDP tunnel offload. GSO packet are almost always CHECKSUM_PARTIAL. USO packets may be CHECKSUM_NONE since commit 10154dbded6d6 ("udp: Allow GSO transmit from devices with no checksum offload"), but then still these fields are initialized correctly in udp4_hwcsum/udp6_hwcsum_outgoing. So no need to test for ip_summed == CHECKSUM_PARTIAL first. This revises an existing fix mentioned in the Fixes tag, which broke small packets with GSO offload, as detected by kselftests. Link: https://syzkaller.appspot.com/bug?extid=e1db31216c789f552871 Link: https://lore.kernel.org/netdev/20240723223109.2196886-1-kuba@kernel.org Fixes: e269d79c7d35 ("net: missing check virtio") Cc: stable@vger.kernel.org Signed-off-by: Willem de Bruijn Link: https://patch.msgid.link/20240729201108.1615114-1-willemdebruijn.kernel@gmail.com Signed-off-by: Jakub Kicinski --- include/linux/virtio_net.h | 16 +++++----------- 1 file changed, 5 insertions(+), 11 deletions(-) (limited to 'include/linux') diff --git a/include/linux/virtio_net.h b/include/linux/virtio_net.h index d1d7825318c3..6c395a2600e8 100644 --- a/include/linux/virtio_net.h +++ b/include/linux/virtio_net.h @@ -56,7 +56,6 @@ static inline int virtio_net_hdr_to_skb(struct sk_buff *skb, unsigned int thlen = 0; unsigned int p_off = 0; unsigned int ip_proto; - u64 ret, remainder, gso_size; if (hdr->gso_type != VIRTIO_NET_HDR_GSO_NONE) { switch (hdr->gso_type & ~VIRTIO_NET_HDR_GSO_ECN) { @@ -99,16 +98,6 @@ static inline int virtio_net_hdr_to_skb(struct sk_buff *skb, u32 off = __virtio16_to_cpu(little_endian, hdr->csum_offset); u32 needed = start + max_t(u32, thlen, off + sizeof(__sum16)); - if (hdr->gso_size) { - gso_size = __virtio16_to_cpu(little_endian, hdr->gso_size); - ret = div64_u64_rem(skb->len, gso_size, &remainder); - if (!(ret && (hdr->gso_size > needed) && - ((remainder > needed) || (remainder == 0)))) { - return -EINVAL; - } - skb_shinfo(skb)->tx_flags |= SKBFL_SHARED_FRAG; - } - if (!pskb_may_pull(skb, needed)) return -EINVAL; @@ -182,6 +171,11 @@ retry: if (gso_type != SKB_GSO_UDP_L4) return -EINVAL; break; + case SKB_GSO_TCPV4: + case SKB_GSO_TCPV6: + if (skb->csum_offset != offsetof(struct tcphdr, check)) + return -EINVAL; + break; } /* Kernel has a special handling for GSO_BY_FRAGS. */ -- cgit v1.2.3 From 21b136cc63d2a9ddd60d4699552b69c214b32964 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Tue, 30 Jul 2024 15:44:16 -0700 Subject: minmax: fix up min3() and max3() too David Laight pointed out that we should deal with the min3() and max3() mess too, which still does excessive expansion. And our current macros are actually rather broken. In particular, the macros did this: #define min3(x, y, z) min((typeof(x))min(x, y), z) #define max3(x, y, z) max((typeof(x))max(x, y), z) and that not only is a nested expansion of possibly very complex arguments with all that involves, the typing with that "typeof()" cast is completely wrong. For example, imagine what happens in max3() if 'x' happens to be a 'unsigned char', but 'y' and 'z' are 'unsigned long'. The types are compatible, and there's no warning - but the result is just random garbage. No, I don't think we've ever hit that issue in practice, but since we now have sane infrastructure for doing this right, let's just use it. It fixes any excessive expansion, and also avoids these kinds of broken type issues. Requested-by: David Laight Acked-by: Arnd Bergmann Signed-off-by: Linus Torvalds --- include/linux/minmax.h | 12 ++++++++++-- 1 file changed, 10 insertions(+), 2 deletions(-) (limited to 'include/linux') diff --git a/include/linux/minmax.h b/include/linux/minmax.h index 41da6f85a407..98008dd92153 100644 --- a/include/linux/minmax.h +++ b/include/linux/minmax.h @@ -152,13 +152,20 @@ #define umax(x, y) \ __careful_cmp(max, (x) + 0u + 0ul + 0ull, (y) + 0u + 0ul + 0ull) +#define __careful_op3(op, x, y, z, ux, uy, uz) ({ \ + __auto_type ux = (x); __auto_type uy = (y);__auto_type uz = (z);\ + BUILD_BUG_ON_MSG(!__types_ok3(x,y,z,ux,uy,uz), \ + #op"3("#x", "#y", "#z") signedness error"); \ + __cmp(op, ux, __cmp(op, uy, uz)); }) + /** * min3 - return minimum of three values * @x: first value * @y: second value * @z: third value */ -#define min3(x, y, z) min((typeof(x))min(x, y), z) +#define min3(x, y, z) \ + __careful_op3(min, x, y, z, __UNIQUE_ID(x_), __UNIQUE_ID(y_), __UNIQUE_ID(z_)) /** * max3 - return maximum of three values @@ -166,7 +173,8 @@ * @y: second value * @z: third value */ -#define max3(x, y, z) max((typeof(x))max(x, y), z) +#define max3(x, y, z) \ + __careful_op3(max, x, y, z, __UNIQUE_ID(x_), __UNIQUE_ID(y_), __UNIQUE_ID(z_)) /** * min_not_zero - return the minimum that is _not_ zero, unless both are zero -- cgit v1.2.3 From 3908ba2e0b2476e2ec13e15967bf6a37e449f2af Mon Sep 17 00:00:00 2001 From: Nick Hu Date: Wed, 17 Jul 2024 11:17:14 +0800 Subject: RISC-V: Enable the IPI before workqueue_online_cpu() Sometimes the hotplug cpu stalls at the arch_cpu_idle() for a while after workqueue_online_cpu(). When cpu stalls at the idle loop, the reschedule IPI is pending. However the enable bit is not enabled yet so the cpu stalls at WFI until watchdog timeout. Therefore enable the IPI before the workqueue_online_cpu() to fix the issue. Fixes: 63c5484e7495 ("workqueue: Add multiple affinity scopes and interface to select them") Signed-off-by: Nick Hu Reviewed-by: Anup Patel Link: https://lore.kernel.org/r/20240717031714.1946036-1-nick.hu@sifive.com Signed-off-by: Palmer Dabbelt --- include/linux/cpuhotplug.h | 1 + 1 file changed, 1 insertion(+) (limited to 'include/linux') diff --git a/include/linux/cpuhotplug.h b/include/linux/cpuhotplug.h index 51ba681b915a..e30d93b807d5 100644 --- a/include/linux/cpuhotplug.h +++ b/include/linux/cpuhotplug.h @@ -148,6 +148,7 @@ enum cpuhp_state { CPUHP_AP_IRQ_LOONGARCH_STARTING, CPUHP_AP_IRQ_SIFIVE_PLIC_STARTING, CPUHP_AP_IRQ_RISCV_IMSIC_STARTING, + CPUHP_AP_IRQ_RISCV_SBI_IPI_STARTING, CPUHP_AP_ARM_MVEBU_COHERENCY, CPUHP_AP_PERF_X86_AMD_UNCORE_STARTING, CPUHP_AP_PERF_X86_STARTING, -- cgit v1.2.3 From b88f55389ad27f05ed84af9e1026aa64dbfabc9a Mon Sep 17 00:00:00 2001 From: Tetsuo Handa Date: Sun, 4 Aug 2024 18:48:10 +0900 Subject: profiling: remove profile=sleep support The kernel sleep profile is no longer working due to a recursive locking bug introduced by commit 42a20f86dc19 ("sched: Add wrapper for get_wchan() to keep task blocked") Booting with the 'profile=sleep' kernel command line option added or executing # echo -n sleep > /sys/kernel/profiling after boot causes the system to lock up. Lockdep reports kthreadd/3 is trying to acquire lock: ffff93ac82e08d58 (&p->pi_lock){....}-{2:2}, at: get_wchan+0x32/0x70 but task is already holding lock: ffff93ac82e08d58 (&p->pi_lock){....}-{2:2}, at: try_to_wake_up+0x53/0x370 with the call trace being lock_acquire+0xc8/0x2f0 get_wchan+0x32/0x70 __update_stats_enqueue_sleeper+0x151/0x430 enqueue_entity+0x4b0/0x520 enqueue_task_fair+0x92/0x6b0 ttwu_do_activate+0x73/0x140 try_to_wake_up+0x213/0x370 swake_up_locked+0x20/0x50 complete+0x2f/0x40 kthread+0xfb/0x180 However, since nobody noticed this regression for more than two years, let's remove 'profile=sleep' support based on the assumption that nobody needs this functionality. Fixes: 42a20f86dc19 ("sched: Add wrapper for get_wchan() to keep task blocked") Cc: stable@vger.kernel.org # v5.16+ Signed-off-by: Tetsuo Handa Signed-off-by: Linus Torvalds --- include/linux/profile.h | 1 - 1 file changed, 1 deletion(-) (limited to 'include/linux') diff --git a/include/linux/profile.h b/include/linux/profile.h index 2fb487f61d12..3f53cdb0c27c 100644 --- a/include/linux/profile.h +++ b/include/linux/profile.h @@ -10,7 +10,6 @@ #define CPU_PROFILING 1 #define SCHED_PROFILING 2 -#define SLEEP_PROFILING 3 #define KVM_PROFILING 4 struct proc_dir_entry; -- cgit v1.2.3