| created | 2019-04-30T17:49:41Z |
|---|---|
| begin | 2019-04-20T00:00:00Z |
| end | 2019-04-21T00:00:00Z |
| path | src/sys |
| commits | 13 |
| date | 2019-04-20T03:10:53Z | |||
|---|---|---|---|---|
| author | jsg | |||
| files | src/sys/dev/pci/drm/radeon/radeon_kms.c | log | diff | annotate |
| message |
Adapt radeondrm_detach_kms() to struct drm_device being split from drm softc. Avoids uvm_fault() when firmware is missing and radeondrm is forcibly detached. Reported by Mihai Popescu on misc@ |
|||
| date | 2019-04-20T08:28:59Z | |||
|---|---|---|---|---|
| author | anton | |||
| files | src/sys/kern/vfs_lockf.c | log | diff | annotate |
| src/sys/sys/lockf.h | log | diff | annotate | |
| message |
Move lockf structures from header to implementation since external users only need a lockf_state pointer by now. ok mpi@ visa@ |
|||
| date | 2019-04-20T08:44:08Z | |||
|---|---|---|---|---|
| author | jsg | |||
| files | src/sys/dev/pci/drm/ttm/ttm_bo.c | log | diff | annotate |
| src/sys/dev/pci/drm/ttm/ttm_memory.c | log | diff | annotate | |
| message |
drm/ttm: Fix bo_global and mem_global kfree error From Trigger Huang b10cc08374728ea79555a1cd98f962b0f942e942 in linux 4.19.y/4.19.36 30f33126feca0fe16df9e9302ffc28a953e2eb37 in mainline linux |
|||
| date | 2019-04-20T11:30:11Z | |||
|---|---|---|---|---|
| author | bluhm | |||
| files | src/sys/netinet6/raw_ip6.c | log | diff | annotate |
| message |
Statistics of "netstat -s -f inet6 -p rip6" did not work. In rip6_sysctl_rip6stat() copy out rip6counters, not ip6counters. OK deraadt@ claudio@ |
|||
| date | 2019-04-20T14:13:11Z | |||
|---|---|---|---|---|
| author | anton | |||
| files | src/sys/kern/vfs_lockf.c | log | diff | annotate |
| message | print locked range in decimal in debug routines | |||
| date | 2019-04-20T15:47:23Z | |||
|---|---|---|---|---|
| author | deraadt | |||
| files | src/sys/conf/param.c | log | diff | annotate |
| message |
as discussed with beck, crank dma-range bufcache to a high number (he suggested 90 but I prefer 80). This is so we learn the downside from user reports. |
|||
| date | 2019-04-20T22:40:13Z | |||
|---|---|---|---|---|
| author | deraadt | |||
| files | src/sys/dev/fdt/acrtc.c | log | diff | annotate |
| message |
armv7 RAMDISK is now compiled with -Oz (just to be like other ramdisks), so clang's rule about "static inline" comes into play. ok patrick |
|||
| date | 2019-04-20T22:59:03Z | |||
|---|---|---|---|---|
| author | deraadt | |||
| files | src/sys/arch/amd64/stand/boot/Makefile | log | diff | annotate |
| src/sys/arch/amd64/stand/cdboot/Makefile | log | diff | annotate | |
| src/sys/arch/amd64/stand/efiboot/Makefile.common | log | diff | annotate | |
| src/sys/arch/amd64/stand/pxeboot/Makefile | log | diff | annotate | |
| src/sys/arch/hppa/stand/boot/Makefile | log | diff | annotate | |
| src/sys/arch/i386/stand/boot/Makefile | log | diff | annotate | |
| src/sys/arch/i386/stand/cdboot/Makefile | log | diff | annotate | |
| src/sys/arch/i386/stand/pxeboot/Makefile | log | diff | annotate | |
| src/sys/arch/loongson/stand/libsa/Makefile | log | diff | annotate | |
| src/sys/arch/luna88k/stand/boot/Makefile | log | diff | annotate | |
| src/sys/arch/macppc/stand/boot.mac/Makefile | log | diff | annotate | |
| message |
libsa's memcpy() is actually memmove(). make a proper memmove(), and give memcpy() correct behaviour. This also brings the bcopy() macro into line. |
|||
| date | 2019-04-20T22:59:04Z | |||
|---|---|---|---|---|
| author | deraadt | |||
| files | src/sys/arch/macppc/stand/ofwboot/Makefile | log | diff | annotate |
| src/sys/arch/octeon/stand/libsa/Makefile | log | diff | annotate | |
| src/sys/arch/sgi/stand/boot/Makefile | log | diff | annotate | |
| src/sys/arch/sgi/stand/libsa/Makefile | log | diff | annotate | |
| src/sys/arch/sparc64/stand/libsa/Makefile | log | diff | annotate | |
| src/sys/lib/libsa/Makefile | log | diff | annotate | |
| src/sys/lib/libsa/memcpy.c | log | diff | annotate | |
| src/sys/lib/libsa/memmove.c | log | diff | annotate | |
| src/sys/lib/libsa/stand.h | log | diff | annotate | |
| message |
libsa's memcpy() is actually memmove(). make a proper memmove(), and give memcpy() correct behaviour. This also brings the bcopy() macro into line. |
|||
| date | 2019-04-20T23:00:07Z | |||
|---|---|---|---|---|
| author | deraadt | |||
| files | src/sys/arch/amd64/stand/libsa/exec_i386.c | log | diff | annotate |
| message |
When copying the EFI-stored kernel to the correct operating location, use memmove. We don't know whether EFI allocated overlapping memory. |
|||
| date | 2019-04-20T23:10:00Z | |||
|---|---|---|---|---|
| author | deraadt | |||
| files | src/sys/sys/exec_elf.h | log | diff | annotate |
| message |
The conversion of rdsetroot to -lelf failed on alpha: -lelf thought ELF SHDR was 80 bytes in size, rather than 64 as listed in the ELF header. In Sep 2001 when ELF was being integrated into the tree, two of the ELF object types (and two more via #define) were given different (incorrect) sizes, and hid behind an #ifdef __alpha__ all this time. -lelf constructs the SHDR object by accumulating sizes of types, so this was finally exposed. A review of the tree shows no other consequences, so we can fix this now. |
|||
| date | 2019-04-20T23:11:20Z | |||
|---|---|---|---|---|
| author | deraadt | |||
| files | src/sys/kern/exec_elf.c | log | diff | annotate |
| message |
#define ELFROUNDSIZE 4 /* XXX Should it be sizeof(Elf_Word)? */ Now that alpha is fixed, we can use sizeof(). |
|||
| date | 2019-04-20T23:44:36Z | |||
|---|---|---|---|---|
| author | kn | |||
| files | src/sys/arch/amd64/amd64/vmm.c | log | diff | annotate |
| message |
Revert enabling VMM_DEBUG which slipped through the previous commit. Fixes panic on `vmctl start foo'. OK tedu deraadt |
|||