x86: unify vmlinux.lds
Get series with b4
b4 am [email protected]Versions newest first
- v1 Latest13 patches 28 repliesIn review
Patches in v1 No applies recorded yet
| # | Subject and files | Size | Replies | Applied |
|---|---|---|---|---|
| 1/13 | x86: beautify vmlinux_32.lds.Sarch/ | +200 | 1 reply | |
| 2/13 | x86, vmlinux.lds: unify header/footerarch/ arch/ arch/ | +72 | 7 replies | |
| 3/13 | x86, vmlinux.lds: unify PHDRSarch/ arch/ arch/ | +13 | 1 reply | |
| 4/13 | x86, vmlinux.lds: unify start/end of SECTIONSarch/ arch/ arch/ | +14 | 1 reply | |
| 5/13 | x86, vmlinux.lds: unify .text output sectionsarch/ arch/ arch/ +1 more | +31 | 1 reply | |
| 6/13 | x86, vmlinux.lds: unify exceptiontablearch/ arch/ arch/ | +10 | 1 reply | |
| 7/13 | x86, vmlinux.lds: unify data output sectionsarch/ arch/ arch/ | +55 | 1 reply | |
| 8/13 | x86, vmlinux.lds: move vsyscall output sectionsarch/ arch/ | +71 | 1 reply | |
| 9/13 | x86, vmlinux.lds: unify first part of initdataarch/ arch/ arch/ | +61 | 1 reply | |
| 10/13 | x86, vmlinux.lds: unify parainstructionsarch/ arch/ arch/ | +18 | 1 reply | |
| 11/13 | x86, vmlinux.lds: unify .exit.* and .init.ramfsarch/ arch/ arch/ | +20 | 1 reply | |
| 12/13 | x86, vmlinux.lds: unify percpuarch/ arch/ arch/ | +30 | 1 reply | |
| 13/13 | x86, vmlinux.lds: unify remaining partsarch/ arch/ arch/ | +27 | 1 reply |
Reviews trailers per patch
1 of 13 patches has a Reviewed-by or Acked-by.
| Patch | Sam Ravnborg |
|---|---|
| 1/13 | |
| 2/13 | A-b |
| 3/13 | |
| 4/13 | |
| 5/13 | |
| 6/13 | |
| 7/13 | |
| 8/13 | |
| 9/13 | |
| 10/13 | |
| 11/13 | |
| 12/13 | |
| 13/13 |
Trailers given in replies are filled; those carried in the patches from earlier versions are outlined with dashes.
Cover letter [PATCH/RFT 0/13] x86: unify vmlinux.lds
Following patchset is on top of the x86/kbuild branch of tip. It unifies vmlinux.lds so we end up with a single file.
The patchset has been build tested on 32bit and 64 bit but has not been boot tested. (Lacking time/resources atm).
The steps were made minimal to make it simpler to track regressions using bisect (if any) and to make review easier.
The unified vmlinux.lds shows some questionable differences between 32 bit and 64 bit.
o 64 bit uses PHDRS more extensively than 32 bit. Could they be the same?
o _stext does not cover all text for 32 bit - a bug? For 64b bit it does. It is only the .code16 wakeup stuff that is not covered but anyway.
o _edata covers much more on 32 bit
o The nosave stuff differs (but that is due to the PHDRS stuff anyway)
o Different alignmnet requirements in several spots
o All the stuff added to support relocable kernels
Please test/review.
This patchset will conflict with Tim's work but the conflicts should be simple to fix. Lets tackle that later.
Sam
arch/x86/kernel/head_32.S | 7 - arch/x86/kernel/vmlinux.lds.S | 421 +++++++++++++++++++++++++++++++++++++- arch/x86/kernel/vmlinux_32.lds.S | 229 --------------------- arch/x86/kernel/vmlinux_64.lds.S | 336 ------------------------------ 4 files changed, 419 insertions(+), 574 deletions(-)
Sam Ravnborg (13):
x86: beautify vmlinux_32.lds.S
x86, vmlinux.lds: unify header/footer x86, vmlinux.lds: unify PHDRS
x86, vmlinux.lds: unify start/end of SECTIONS x86, vmlinux.lds: unify .text output sections x86, vmlinux.lds: unify exceptiontable x86, vmlinux.lds: unify data output sections x86, vmlinux.lds: move vsyscall output sections x86, vmlinux.lds: unify first part of initdata x86, vmlinux.lds: unify parainstructions x86, vmlinux.lds: unify .exit.* and .init.ramfs x86, vmlinux.lds: unify percpu
x86, vmlinux.lds: unify remaining parts