Skip to contentSkip to conversation

e1000: Fix PCI memory addressing and fix for receive data polling

RFC PATCH New Ralf Rösch 2 participants 2 messages Jan 26, 2012

lore
Share & tools
lore

Ralf Rösch

author
#0 lore raw

RFC PATCH 1/1 e1000: Fix PCI memory addressing and fix for receive data polling

Sorry, you will find now my fixes in attachment

Hi all,

I had some problem when using your e1000-code on a MIPS target. Attached you will find two fixes which I would like to propose for discussion.

- Fix 1: this has the same reason as in commit
f81ecb5d3300bf92d17302d3712f30585c182da9

- Fix 2: has to do with "C Language Operator Precedence".
I would say the expression taken by the ! operator has precedence before the & operator.
So any bit in rd->status other than E1000_RXD_STAT_DD stops correct detection of data reception.
(What i do not know is, if it is possible that any other bit than E1000_RXD_STAT_DD can get active)

Please leave me in CC, because i'm not subscribed to this list.

Thanks a lot for providing u-boot.

Best regards
Ralf

Signature · 18 lines
Roesch & Walter___________________________________________
Industrie-Elektronik GmbH * Tel.: +49-7824 / 6628-0
W?rtelweg 2b/c            * Fax:  +49-7824 / 6628-29
D-77963 Schwanau          * mailto:ralf.roesch at rw-gmbh.de
Germany                   * WWW: http://www.rw-gmbh.de
Amtsgericht Freiburg i.Br. HRB 391345
Gesch?ftsf?hrer: Dipl.Ing.(FH) Ralf R?sch, Dipl.Ing.(FH) Martin Walter

GnuPG key: 52ECD70F 2010-09-04 [expires: 2012-12-31]
Fingerprint: 8415 9113 5F05 D579 6685  D5AD 5CE7 5429 52EC D70F

-------------- next part --------------
A non-text attachment was scrubbed...
Name: e1000.patch
Type: text/x-diff
Size: 828 bytes
Desc: not available
URL: <http://lists.denx.de/pipermail/u-boot/attachments/20120126/9a43b66f/attachment.patch>
#1 lore raw

RFC PATCH 1/1 e1000: Fix PCI memory addressing and fix for receive data polling

Quoted 3 lines · Ralf R?sch

Sorry, you will find now my fixes in attachment

Hi all,

Hi Ralf,

Quoted 7 lines

I had some problem when using your e1000-code on a MIPS target. Attached you will find two fixes which I would like to propose for discussion.

- Fix 1: this has the same reason as in commit
f81ecb5d3300bf92d17302d3712f30585c182da9

Anyway, can you also describe *which* bug / behavior have you found ?

Quoted 7 lines

- Fix 2: has to do with "C Language Operator Precedence".
I would say the expression taken by the ! operator has precedence before the & operator.
So any bit in rd->status other than E1000_RXD_STAT_DD stops correct detection of data reception.
(What i do not know is, if it is possible that any other bit than E1000_RXD_STAT_DD can get active)

This seems to me a bug. However, you must send right patches for review / discussion. You find two issues, and you must send one patch for each issue that will be fixed. Please have a look at:

http://www.denx.de/wiki/U-Boot/Patches

It describes how to submit patches in the right format.

Best regards,
Stefano Babic

Signature · 5 lines
=====================================================================
DENX Software Engineering GmbH,     MD: Wolfgang Denk & Detlev Zundel
HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany
Phone: +49-8142-66989-0 Fax: +49-8142-66989-80  Email: office at denx.de
=====================================================================