6.6-stable review patch. If anyone has any objections, please let me know. ------------------ From: Terry Bowman [ Upstream commit c268f949e219f9e179558e836f457f6c5fbec416 ] The CXL r4.0 8.2.4.17.7 RAS Capability Structure has total length 0x58 bytes (CXL_RAS_CAPABILITY_LENGTH); the Header Log occupies the trailing 64 bytes at offset 0x18. CXL_HEADERLOG_SIZE was defined as SZ_512, eight times the actual on-device size. header_log_copy() reads CXL_HEADERLOG_SIZE_U32 (128) dwords from the RAS capability iomap, overrunning the 88-byte mapping by 448 bytes. The cxl_aer_uncorrectable_error trace event memcpy()s CXL_HEADERLOG_SIZE (512) bytes from its source. For the CPER caller the source is struct cxl_ras_capability_regs::header_log[16] (64 bytes) embedded in a stack-local cxl_cper_prot_err_work_data, so the memcpy reads 448 bytes of kernel stack into the trace event ring buffer where userspace can read it via tracefs. Set CXL_HEADERLOG_SIZE to 64 and derive CXL_HEADERLOG_SIZE_U32 from it, bringing all iomap readers into agreement on 16 dwords. Userspace tools such as rasdaemon have grown a dependency on the buggy 512-byte (128 u32) header_log layout in the cxl_aer_uncorrectable_error trace event. Add CXL_HEADERLOG_TRACE_SIZE_U32 = 128 and use it for the trace event __array and its memcpy to preserve that ABI. Both callers now pass a zero-filled u32[CXL_HEADERLOG_TRACE_SIZE_U32] staging buffer with only the first CXL_HEADERLOG_SIZE_U32 (16) entries populated from hardware; the remaining 112 u32s are zero-padded, keeping the 512-byte trace ring buffer layout intact. [ dj: Replaced 64 with SZ_64 per RichardC ] Fixes: 36f257e3b0ba ("acpi/ghes, cxl/pci: Process CXL CPER Protocol Errors") Fixes: 2905cb5236cb ("cxl/pci: Add (hopeful) error handling support") Cc: stable@vger.kernel.org Reported-by: Sashiko Signed-off-by: Terry Bowman Reviewed-by: Alison Schofield Reviewed-by: Dave Jiang Reviewed-by: Ben Cheatham Reviewed-by: Richard Cheng Link: https://patch.msgid.link/20260605180610.2249458-1-terry.bowman@amd.com Signed-off-by: Dave Jiang Signed-off-by: Sasha Levin --- drivers/cxl/core/pci.c | 10 +++++++--- drivers/cxl/core/trace.h | 12 ++++++++---- drivers/cxl/cxl.h | 14 ++++++++++++-- 3 files changed, 27 insertions(+), 9 deletions(-) diff --git a/drivers/cxl/core/pci.c b/drivers/cxl/core/pci.c index 6edfd05466737..e89d4a9827155 100644 --- a/drivers/cxl/core/pci.c +++ b/drivers/cxl/core/pci.c @@ -11,6 +11,10 @@ #include "core.h" #include "trace.h" +/* Check that UCE header definition is maintained to keep ABI intact */ +static_assert(CXL_HEADERLOG_TRACE_SIZE_U32 == 128, + "rasdaemon ABI requires exactly 128 u32s"); + /** * DOC: cxl core pci * @@ -663,12 +667,12 @@ static void header_log_copy(struct cxl_dev_state *cxlds, u32 *log) { void __iomem *addr; u32 *log_addr; - int i, log_u32_size = CXL_HEADERLOG_SIZE / sizeof(u32); + int i; addr = cxlds->regs.ras + CXL_RAS_HEADER_LOG_OFFSET; log_addr = log; - for (i = 0; i < log_u32_size; i++) { + for (i = 0; i < CXL_HEADERLOG_SIZE_U32; i++) { *log_addr = readl(addr); log_addr++; addr += sizeof(u32); @@ -681,7 +685,7 @@ static void header_log_copy(struct cxl_dev_state *cxlds, u32 *log) */ static bool cxl_report_and_clear(struct cxl_dev_state *cxlds) { - u32 hl[CXL_HEADERLOG_SIZE_U32]; + u32 hl[CXL_HEADERLOG_TRACE_SIZE_U32] = {}; void __iomem *addr; u32 status; u32 fe; diff --git a/drivers/cxl/core/trace.h b/drivers/cxl/core/trace.h index 1e1e50d39bf63..54b72f943744b 100644 --- a/drivers/cxl/core/trace.h +++ b/drivers/cxl/core/trace.h @@ -57,7 +57,7 @@ TRACE_EVENT(cxl_aer_uncorrectable_error, __field(u64, serial) __field(u32, status) __field(u32, first_error) - __array(u32, header_log, CXL_HEADERLOG_SIZE_U32) + __array(u32, header_log, CXL_HEADERLOG_TRACE_SIZE_U32) ), TP_fast_assign( __assign_str(memdev, dev_name(&cxlmd->dev)); @@ -66,10 +66,14 @@ TRACE_EVENT(cxl_aer_uncorrectable_error, __entry->status = status; __entry->first_error = fe; /* - * Embed the 512B headerlog data for user app retrieval and - * parsing, but no need to print this in the trace buffer. + * Embed headerlog data for user app retrieval and parsing, + * but no need to print in the trace buffer. Only + * CXL_HEADERLOG_SIZE_U32 (16) dwords are hardware data; + * the remaining entries preserve the 512-byte ABI layout + * rasdaemon depends on and are zero-filled by the caller. */ - memcpy(__entry->header_log, hl, CXL_HEADERLOG_SIZE); + memcpy(__entry->header_log, hl, + CXL_HEADERLOG_TRACE_SIZE_U32 * sizeof(u32)); ), TP_printk("memdev=%s host=%s serial=%lld: status: '%s' first_error: '%s'", __get_str(memdev), __get_str(host), __entry->serial, diff --git a/drivers/cxl/cxl.h b/drivers/cxl/cxl.h index f3260ca17522a..9dcc8a4d5128a 100644 --- a/drivers/cxl/cxl.h +++ b/drivers/cxl/cxl.h @@ -143,8 +143,18 @@ static inline int ways_to_eiw(unsigned int ways, u8 *eiw) #define CXL_RAS_CAP_CONTROL_FE_MASK GENMASK(5, 0) #define CXL_RAS_HEADER_LOG_OFFSET 0x18 #define CXL_RAS_CAPABILITY_LENGTH 0x58 -#define CXL_HEADERLOG_SIZE SZ_512 -#define CXL_HEADERLOG_SIZE_U32 SZ_512 / sizeof(u32) +#define CXL_HEADERLOG_SIZE SZ_64 +#define CXL_HEADERLOG_SIZE_U32 (CXL_HEADERLOG_SIZE / sizeof(u32)) + +/* + * The RAS UCE trace event header array was originally sized at SZ_512/sizeof(u32) + * = 128 u32s due to a bug. Userspace tools (rasdaemon) have grown a dependency + * on that 512-byte layout. Keep the trace array at 128 u32s to preserve the + * ABI; only CXL_HEADERLOG_SIZE_U32 (16) dwords are valid hardware data, the + * remainder are zero-filled. + */ +#define CXL_HEADERLOG_TRACE_SIZE SZ_512 +#define CXL_HEADERLOG_TRACE_SIZE_U32 (CXL_HEADERLOG_TRACE_SIZE / sizeof(u32)) /* CXL 2.0 8.2.8.1 Device Capabilities Array Register */ #define CXLDEV_CAP_ARRAY_OFFSET 0x0 -- 2.53.0