Re: [PATCH 12/19] RISC-V: riscv_set_arch() can fail
Project / Subsystem
binutils / risc-v
Date
2026-04-29
Proposer
Jiawei <jiawei@iscas.ac.cn>
Source type
public_inbox
Consensus
Ready to Land
Sentiment
—/10
Technical tradeoffs
- • Using a default architecture string might not always be the desired behavior, but it avoids a crash.
- • The assembler will proceed with a default arch if the expected arch string is missing.
All attributes
- project
- binutils
- subsystem
- risc-v
- patch_id
- —
- discussion_id
- 9e1a5fd2-e547-4bbb-be77-c97e115e40bf@iscas.ac.cn
- source_type
- public_inbox
- title
- Re: [PATCH 12/19] RISC-V: riscv_set_arch() can fail
- headline
- RISC-V: riscv_set_arch() can fail
- tldr
- Avoid passing NULL pointers to functions when ``riscv_set_arch()`` fails to set architecture string.
- proposer
- Jiawei <jiawei@iscas.ac.cn>
- consensus
- Ready to Land
- outcome
- ready
- sentiment_score
- —
- technical_tradeoffs
-
- • Using a default architecture string might not always be the desired behavior, but it avoids a crash.
- • The assembler will proceed with a default arch if the expected arch string is missing.
- series_id
- binutils:risc-v: riscv_set_arch() can fail
- series_role
- reply
- series_parts
- []
- tags
-
- • risc-v
- • error handling
- • null pointer
- bugzilla_url
- —
- date
- 2026-04-29T00:00:00.000Z
Re: [PATCH 12/19] RISC-V: riscv_set_arch() can fail
This patch fixes a potential crash in riscv_set_arch() when the architecture string is empty. In such cases, the code now avoids passing NULL pointers to subsequent functions by using a default architecture string (DEFAULT_RISCV_ARCH_WITH_EXT). This prevents crashes when the frontend fails to provide a valid architecture string.