On 2 February 2017 at 18:36, Ard Biesheuvel <ard.biesheuvel@xxxxxxxxxx> wrote: > On 2 February 2017 at 18:30, Will Deacon <will.deacon@xxxxxxx> wrote: >> Hi Ard, >> >> On Thu, Feb 02, 2017 at 05:33:19PM +0000, Ard Biesheuvel wrote: >>> When building with debugging symbols, take the absolute path to the >>> vmlinux binary and add it to the special PE/COFF debug table entry. >>> This allows a debug EFI build to find the vmlinux binary, which is >>> very helpful in debugging, given that the offset where the Image is >>> first loaded by EFI is highly unpredictable. >>> >>> On implementations of UEFI that choose to implement it, this >>> information is exposed via the EFI debug support table, which is a UEFI >>> configuration table that is accessible both by the firmware at boot time >>> and by the OS at runtime, and lists all PE/COFF images loaded by the >>> system. >>> >>> The format of the NB10 Codeview entry is based on the definition used >>> by EDK2, which is our primary reference when it comes to the use of >>> PE/COFF in the context of UEFI firmware. >>> >>> Signed-off-by: Ard Biesheuvel <ard.biesheuvel@xxxxxxxxxx> >>> --- >>> v4: add a separate DEBUG_EFI debug option so that this feature is disabled >>> by default >>> dropped 1/2 which has been queued already in the EFI tree >>> use ASCII rather than an opaque constant for "NB10" signature >>> >>> arch/arm64/Kconfig.debug | 8 ++++ >>> arch/arm64/kernel/Makefile | 4 ++ >>> arch/arm64/kernel/head.S | 47 +++++++++++++++++++- >>> 3 files changed, 58 insertions(+), 1 deletion(-) >>> >>> diff --git a/arch/arm64/Kconfig.debug b/arch/arm64/Kconfig.debug >>> index d1ebd46872fd..f7f38b1aab14 100644 >>> --- a/arch/arm64/Kconfig.debug >>> +++ b/arch/arm64/Kconfig.debug >>> @@ -95,6 +95,14 @@ config DEBUG_ALIGN_RODATA >>> >>> If in doubt, say N. >>> >>> +config DEBUG_EFI >>> + depends on EFI && DEBUG_INFO >>> + bool "UEFI debugging" >>> + help >>> + Enable this option to include EFI specific debugging features into >>> + the kernel that are only useful when using a debug build of the >>> + UEFI firmware >>> + >> >> I don't much like embedding the absolute host build path in the kernel >> image, but if that's what the tools expect then so be it. >> >>> source "drivers/hwtracing/coresight/Kconfig" >>> >>> endmenu >>> diff --git a/arch/arm64/kernel/Makefile b/arch/arm64/kernel/Makefile >>> index 7d66bbaafc0c..2600c60337ac 100644 >>> --- a/arch/arm64/kernel/Makefile >>> +++ b/arch/arm64/kernel/Makefile >>> @@ -55,3 +55,7 @@ obj-y += $(arm64-obj-y) vdso/ probes/ >>> obj-m += $(arm64-obj-m) >>> head-y := head.o >>> extra-y += $(head-y) vmlinux.lds >>> + >>> +ifeq ($(CONFIG_DEBUG_EFI),y) >>> +AFLAGS_head.o += -DVMLINUX_PATH="\"$(shell readlink -f $(objtree)/vmlinux)\"" >>> +endif >> >> Can we use the Make realpath directive instead of calling out to the shell? >> > > Yep. This works as expected: > > diff --git a/arch/arm64/kernel/Makefile b/arch/arm64/kernel/Makefile > index 2600c60337ac..1606c6b2a280 100644 > --- a/arch/arm64/kernel/Makefile > +++ b/arch/arm64/kernel/Makefile > @@ -57,5 +57,5 @@ head-y := head.o > extra-y += $(head-y) vmlinux.lds > > ifeq ($(CONFIG_DEBUG_EFI),y) > -AFLAGS_head.o += -DVMLINUX_PATH="\"$(shell readlink -f $(objtree)/vmlinux)\"" > +AFLAGS_head.o += -DVMLINUX_PATH="\"$(realpath $(objtree)/vmlinux)\"" > endif > > >> Do you want me to merge this via arm64? >> > > Yes, please Should I respin this? -- To unsubscribe from this list: send the line "unsubscribe linux-efi" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
