cmd/compile: linux/mips fails to make.bash with "internal compiler error: panic: runtime error: index out of range [0] with length 0"

[dmitshur]

From https://build.golang.org/log/5e58860b238976b022020c1cb9b3537a20adc863:

Building Go cmd/dist using /tmp/gobuilder-mips64/go1.4/go. (go1.22.6 linux/mips64)
Building Go toolchain1 using /tmp/gobuilder-mips64/go1.4/go.
Building Go bootstrap cmd/go (go_bootstrap) using Go toolchain1.
Building Go toolchain2 using go_bootstrap and Go toolchain1.
Building Go toolchain3 using go_bootstrap and Go toolchain2.
# internal/unsafeheader
<unknown line number>: internal compiler error: panic: runtime error: index out of range [0] with length 0

[...]

This isn't happening on other ports, but does happen on GOOS=linux GOARCH=mips. Also note that without a LUCI builder (issue #67303), watchflakes doesn't catch and report persistent failures such as this one.

https://build.golang.org/log/e098e9f4dd5cedc9afd00c53a1ead71b3ab66125 was the last passing commit, so this seems to be related to CL 594597 for #54766.

CC @golang/mips.

[gabyhelp]

Related Issues

• cmd/compile: index of out of range compiling internal/obj/arm64 on linux-mipsle #19641 (closed)
• cmd/go: go1.11.4 toolchain3 build fail with "slice bounds out of range" on 32-bit MIPS on Debian buildd #29402 (closed)
• cmd/compile: building causes a panic: runtime error: index out of range [-1] #47799 (closed)
• cmd/link/internal/loader: panic: runtime error: index out of range [1232602353] with length 116 #69214 (closed)
• build: linux/mipsle failure -- memory corruption? #19396 (closed)
• cmd/link: index out of range when compiling plugin #17140 (closed)
• cmd/link: panic: runtime error: index out of range #10718 (closed)
• cmd/compile: ICE: "runtime error: index out of range [0] with length 0" due to generic code #52162 (closed)
• cmd/compile: internal compiler error: panic during lower (integer divide by zero) #43099 (closed)
• build: crashes during compilation on mips64le hw #16848 (closed)

_{(Emoji vote if this was helpful or unhelpful; more detailed feedback welcome in this discussion.)}

[prattmic]

mipsle is OK, so this is probably a big endian problem. I thought I handled this with ppc64, but perhaps this is 32-bit big endian specific.